470 ADAPTATIONS TO PHOTIC QUALITY 



Suppose the woodchuck has learned perfectly to go always to the blue 

 member of a blue-red pair of stimuli. You may now begin changing the 

 red stimulus in brightness, to look for a point at which the two stimuli 

 are equally bright to the animal. There is sure to be such a value of in- 

 tensity of the red light, but if you are fortunate you will not identify it. 

 If the animal has been discriminating the colors as such, he will continue 

 to go to the blue no matter how the red may be altered bit by bit, trial by 

 trial or day by day, up or down, in intensity. If on the contrary he has 

 been going to the blue only because it was brighter, say, than the red for 

 his eye (it matters not that the two may originally have seemed equally 

 bright to your own eyes) , then at some intensity value of the red stimulus 

 he will break down and make chance scores — that is, go as often to one 

 stimulus as to the other. On the face of things this will indicate that he 

 is totally color-blind; but he may only have been paying more attention 

 to the greater brightness of the blue than to its coloredness. In this case 

 he will soon make the 'blue = food' association once more, even with the 

 red stimulus held constant at the confusing value. But if he continues 

 indefihitely to make chance scores, and goes to the red stimulus when 

 this is brighter for him than the blue (the 'step-wise phenomenon' — 

 always going to the brighter, or less bright, of any two stimuli) , then he 

 surely has no color-discriminatory capacity whatever. 



If the two stimuli first used were not from near the opposite ends of 

 the spectrum, however, the animal might break down at particular inten- 

 sity-values, and still have color vision, of a diflFerent character from the 

 normal human. Yellow and orange, or blue and violet, would probably 

 look alike in hue to an animal with a dichromatic system anything like 

 that of a 'color-blind' human (see pp. 97-9), whereas either member of 

 one pair would always be discriminable from either member of the other 

 pair, since the two pairs are on opposite sides of any possible dichromate 

 neutral point. Where there is any suspicion that the species in hand is 

 dichromatic, a search must be made for a neutral point or band — a 

 spectral region which the animal cannot distinguish from a white light. 

 But it so happens that no vertebrate species (unless it be the cebus mon- 

 key — see p. 516) has yet been found to have a dichromatic system, 

 though unfortunately few investigators have so devised their experiments 

 as to disclose dichromasy even if it were possessed by their particular 

 animals. 



If colored papers or objects are used, it is most convenient to eliminate 

 the possibility of a brightness-basis for the discriminations by training the 



