634 THE EYE IN EVOLUTION 



could be trained in behavioural experiments to distinguish red, yellow, 

 green and blue papers from one another and from grey of equal 

 brightness. With the European ground-squirrel (souslik, Citellus 

 citellus), Kolosvary (1934) concluded that a colour-preference existed 

 for blue. None of these experiments is fully convincing, but it seems 

 that a weak capacity for colour vision may exist in some species of 

 squirrel, while in many individuals it is wholly absent ; even if it is 

 occasionally present, it seems unlikely that it can determine behaviour. 

 Among the carnivores, the earlier workers gave most attention 

 to the dog. The variable results initially obtained are vitiated by 

 absence of the adequate control of intensity (Lubbock, 1888 ; Gates, 

 1895 ; Himstedt and Nagel, 1902 ; Nicolai, 1907 ; Orbeli, 1909 ; 

 Colvin and Burfo^d, 1909 ; Kahscher, 1909) ; while the better con- 

 trolled experiments of Samoiloff and Pheophilaktova (1907) and 

 E. M. Smith (1912) led to the conclusion that hues have little significance 

 for this animal despite its undoubted intelligence and amenability to 

 experimental restraints. Confusion of coloured papers with greys was 

 practically constant although after prolonged training some animals 

 seemed to show some recognition of green. It will be remembered that 

 Orbeli (1909) in Pavlov's laboratory, found similarly inconclusive 

 results on attempting to establish conditioned reflexes to colours in 

 this animal.^ All observers are agreed that colours have no significance 

 whatever for the cat whether attempts at training have been made by 

 coloured papers or spectral lights (de Voss and Ganson, 1915 ; Gregg 

 et al., 1929 ; Gunter, 1952-54 ; Meyer et al, 1954) ; the positive 

 results claimed by Colvin and Burford (1909) and Kalischer (1909) can 

 be explained by inadequate controls and the mistake of equating 

 relative brightness to human standards. A similar criticism applies to 

 the claim of Cole (1907) and Cole and Long (1909) that the raccoon has 

 some degree of colour vision ; Davis (1907) and Gregg and his co- 

 . workers (1929) obtained completely negative results with this animal, 

 as did Miiller (1930) in the marten, Martes, and the pole-cat, Putorius. 

 It would seem, therefore, that with the problematical exception of the 

 dog,'^ all the Carnivores so far tested have proved to be colour-blind 

 or have indicated that colours have no significance for them. If in 

 some dogs some discrimination of hue is possible, the faculty seems to 

 be without importance to the animal and is entirely dominated by 

 sensations of form and brightness. 



UNGULATES which have been investigated have been found to be 

 similar. Cattle — even the fiajhting bulls of Latin Europe and America — 

 are completely colour-bKnd (Kittredge, 1923 ; Stratton, 1923) ; they 



» p. 623. 



" See, however, Schubert (1950) who, while admitting that it has not been shown 

 experimentally that dogs have hue-discrimination analogous to man, insists that these 

 animals are not colour-blind. 



