620 THE EYE IN EVOLUTION 



far as is known, the spectral limits of the vision of all Vertebrates are 

 approximately the same, and include nothing corresponding to the 

 visibility of the ultra-violet to insects.^ 



Within this spectral range the human eye can differentiate many 

 hues, qualities independent of the luminosity ; and to man, colour 

 sensations are highly overlaid with aesthetic values. These, however, 

 must take a subsidiary place in vertebrate evolution ; in the animal 

 hue-discrimination is never developed to a corresponding extent and 

 although in certain species it may have a secondary biological veJue 

 in sexual displays or as a means of concealment and advertisement, it 

 would appear to be essentially a mechanism designed to increase the 

 visual acuity by acting as an adjuvant to the discrimination of bright- 

 ness. Evolved out of the light-sense with a view to obtaining a more 

 critical analysis than could be provided by the appreciation of differ- 

 ences in luminosity alone, there is no legitimate reason to assume that 

 different bands of the spectrum excite in animals the perceptual 

 experiences recognized as colour by us. Moreover, as would be expected 

 from its biological purpose, hue-discrimination is found largely in 

 intensely visual Vertebrates with highly diurnal activities and pro- 

 vided with a cone-rich retina, a fovea and an effective accommodative 

 apparatus ; a colour sense, in fact, is associated with good visual 

 acuity, and that we shall see presently ^ is rare among Vertebrates. 

 When vision is vague and limited largely to an appreciation of luminosity 

 and movement, the refined discrimination provided by colour vision 

 is likely to be of little importance ; and to the nocturnal animal such 

 a faculty would seem to be meaningless. 



The investigation of colour vision in animals has excited much 

 attention since the early work of Graber (1884-85) on fish. Even in 

 human experiments this is a notoriously difficult subject, but in animals 

 the difficulties increase manyfold ; unless the stimulus is presented 

 with the utmost care it is difficult to exclude variants other than hue- 

 discrimination, such as changes in luminosity, in any choice the animal 

 may make. Moreover, the tractability, responsiveness and intelligence 

 required to produce a consistent response are frequently lacking so that 

 in many cases a lack of a colour sense may be presumed when conduct 

 may have been determined by irritability, untrainability or brainlessness. 

 If, for example, a colour sense is not highly developed and does not 

 play a prominent part in the everyday behaviour of the animal, 

 experimental testing will probably involve difficult discriminations 

 comparable to a complicated intelligence test in man in which the 

 subject, unable to grasp completely the point at issue, has to rely on 



1 Owls have been credited with vision in the infra-red, but this is not the case 

 — see p. 630. 

 * p. 637. 



