THE PERCEPTION OF COLOUR 621 



guesses which may often be misleading and are rarely consistent. ^ 

 Because this limitation has not been realized, much of the work on 

 this subject has been scientifically worthless and in the literature many 

 contradictions are to be found. 



The methods employed in the exploration of the colour vision of 

 animals embrace the two classes we have already discussed as being 

 available for the analysis of other visual functions ^ — objective and 

 subjective. 



THE OBJECTIVE METHODS OF APPROACH depend on the observation 

 of a measurable physical phenomenon presumed to be determined by 

 a specificity in the retinal response to different wave-lengths ; they 

 suffer from the weakness that such a differential response does not 

 necessarily imply a conscious appreciation of hue. The luminosity 

 curve for the dark-adapted human eye, for example, shows a differen- 

 tial sensitivity to different wave-lengths and yet does not imply a 

 sensation of colour. Even although more than one retinal mechanism 

 may be stimulated and a physical basis may be shown to exist where- 

 upon colour vision could be based, vision on the perceptual level 

 may nevertheless be achromatic. Indeed, as Pumphrey (1949) sug- 

 gested, it may well be that most animals with highly developed eyes 

 have the fundamental mechanism for mediating colour vision, but it 

 is utilized only by the few to which it is a biologically useful attribute. 



Several such phenomena have been utilized : 



(a) Dermal colour changes. One of tlie earliest arguments employed in 

 ascribing the faculty of colour vision to aninials was the occurrence of changes 

 in colour in the integument or its appendages in response to the environment,^ 

 a study which was first applied on a scientific basis by Karl von Frisch (1912) 

 to fishes ; it seems unlikely that Nature wovild evolve a complicated method 

 of camouflage based on colour if differences in hue were not appreciated by the 

 enemies it was advisable to avoid, while the assumption of brilliant colours as 

 a method of sexual attraction becomes meaningless if the potential mate is 

 unresponsive to the stimulus so elaborately provided. This is true ; but it is to 

 be remembered that many of the colour changes designed to mimic an environ- 

 ment are reflex * and need not enter into consciousness, and even if they are it 

 is conceivable that in some cases the changes appreciated by us as hue may be 

 interjDreted by soine animals in terms of luminosity, providing changes in contrast 

 rather than in quality. This approach is therefore suggestive, particularly in 

 the case of teleostean Fishes and Birds, but cannot be accepted as implying 

 rigid and incontrovertible proof of the existence of true colour vision. Moreover, 

 if it is used at all, the method must be employed only on adequately controlled 

 experiinental trials. 



(6) The pupillary reactions. Observations of the differential effect of wave- 

 bands in the spectruin in the induction of pupillary contractions stem fi'O in the 

 original observation of Sachs (1892-1900) that with lights of ec^ual energy this 



1 See Smith (1912). 2 p. 568. 



3 p. 82. * p. 92. 



