700 THE EYE IN EVOLUTION 



nuclei, with their close anatomical relationship to the posterior longitudinal 

 bundle and the vestibular system, their secondary changes in position correspond- 

 ing to changes in the paths of the optic, vestibular and coordinative reflexes. 

 In the present case, when the eyes are directed frontally but are in a non- 

 converging position, the nasal fibres of one retina and the temporal fibres of the 

 other are stimulated simultaneously by laterally incident light ; these fibres 

 therefore rvm in contiguity in the central nervous tract. Again, with frontally 

 incident light, the image is formed on the temporal sides of the retinae of both 

 eyes. Thus the temporal region of one retina works partly with the opposite 

 temporal region and partly with the opposite nasal region, whereas the nasal 

 regions never work together. Hence the temporal fibres from both sides must 

 also run in contiguity, and therefore there are both direct and crossed (macular) 

 temporal fibres in each tract. 



Associated with the motorial coordination of the eyes the pupillary reactions 

 are interesting. There is a consensual pupillary reaction in the selachian rays 

 and in the pigeon, but so far as is known in all the other lower Vertebrates 

 wherein the pupils react to light — Fishes, Amphibians, Reptiles and Birds — the 

 reaction is unilateral and confined to the stimulated eye (Rochon-Duvigneaud, 

 1943). In the lower Mammals such as the Rodents the same unilaterality 

 obtains ; a faint consensual reaction is seen in Carnivores such as the cat and 

 dog in which the non-decussation of nerve fibres becomes considerable ; while 

 only in Primates wherein a hemi-decussation occurs do the sensory and motorial 

 reactions become fully conjugated and the responses of the two pupils become 

 almost equal when one eye is stimulated. 



SPATIAL JUDGMENTS 



While no systematic research has been devoted to the subject, the 

 visual performance of Vertebrates leaves little doubt that spatial 

 perceptions of some accuracy are a universal attribute of vertebrate 

 vision, probably crude in the uniocular field, often of great accuracy in 

 the binocular field and sometimes of incredible accuracy with bifoveate 

 vision. If we reason from our own subjective impressions — always, it 

 is to be remembered, a most dangerous thing to do — it is probable that 

 in the uniocular field these perceptions are derived from such factors 

 as the retinal size of the images of known objects, overlap of contours, 

 the placement of shadows, aerial perspective and uniocular parallax, 

 often with the help of accommodation. Within the binocular field 

 clues of greater accuracy are provided by the disparity of the retinal 

 images seen by the two eyes and the effort expended in convergence, 

 while in Placentals in which the eyes are coordinated physiological 

 diplopia probably becomes a potent factor in stereopsis for near objects, 

 together with parallactic localization of an object in space. 



That uniocular clues do play a considerable part in spatial 

 perceptions in animals is obvious from the visual judgments formed 

 by many animals with panoramic vision, and is confirmed by several 

 observations. We have already noted the jerky or oscillatory move- 

 ments of the head so constantly seen in many birds ; viewing space 



