THE PERCEPTION OF COLOUR 623 



has been proved to exist which could be used for the differentiation of hues 

 there is no reason why the animal should not have achromatic vision. Moreover 

 in using electrophysiological experiments to interpret the more complex visual 

 mechanisms such as colour vision, which presumably depends on the simul 

 taneous recognition of unlike messages from different optic nerve fibres, con 

 elusions cannot be based on the discharges picked up from the whole retina or 

 optic nerve bv;t only froin the analysis of those derived from individual elements 

 this was not done by the earlier workers. 



(d) Reflex responses. Conditioned reflexes have been employed to elucidate 

 the problem of colour vision, first by Orbeli (1909), in Pavlov's laboratory, who 

 studied the effect of conditioned coloured stimuli on salivary secretion in dogs ; 

 the results were inconclusive and largely negative. More conckisive evidence 

 was obtained by Bull (1935) working on conditioned reflexes established on a 

 basis of wave-discrimination by the blenny, Blennius pholis. 



Other reflex responses have been vitilized in the study of colour vision, 

 such as changes in the respiration rate that occur when some fish are exposed 

 to lights of different colours (Reeves, 1919) or changes in the reflex action of 

 posture when the two eyes are unequally stimulated (Thibault, 1949). All svich 

 methods are of considerable corroboratory value but their results can be 

 translated into terms of sensation only with diffidence. 



(e) The optomotor reaction has been pressed into the service of the exploration 

 of colour vision. Therein, it will be remembered, the animal is faced with a 

 revolving drum with vertical stripes and if these can be differentiated, com- 

 pensatory movements of the eyes occur. Schlieper (1927) reasoned that if a 

 shade of grey were found which elicited no movements when alternated with 

 stripes of a colour, the field must appear hoinogeneous and the animal must 

 therefore be colour-blind to that colovir ; from this negative response he conckided 

 that the fishes and lizards with which he experimented only responded to differ- 

 ences in brightness and not in hue. Others have subsequently exploited the 

 method, particularly Birukow (1937-50) with Amphibians, but again, the presence 

 of a reflex response on a physical level, although suggestive, does not demonstrate 

 the presence of colour appreciation on the physiological level. 



SUBJECTIVE METHODS OF BEHAVIOURAL DISCRIMINATION are mUch 



more satisfying from the physiological jDoint of view than objective 

 responses since they imply the presence of the faculty to differentiate 

 hues as sensations. Unfortunately much of the earlier work on this 

 subject is lacking in adequate control, the principal fault being the 

 failure to appreciate the importance of the elimination of differences 

 in luminosity from the stimulus or, alternatively, the widespread tacit 

 suggestion that the appreciation of luminosity (or of hue) of an animal 

 can be legitimately equated to human sensations or to standards based 

 on equality of energy. The assumption that the appreciation of light 

 or colour by any species of animal resembles that of any other species, 

 including man. rests on inseciu'e evidence. 



The simplest experimental technique is that of colour-preference — the simple 

 observation of whether the animal prefers to go towards one colour before another. 

 This crvide method was first employed by Graber (1884-85) who found that 

 certain teleostean fishes preferentially swam to a light rather than darkness, 

 and to red rather than green rather than blue — the " step -wise phenomenon." 



