326 ADAPTATIONS TO SPACE AND MOTION 



analogy for corresponding retinal points, for if two adjacent fingers be 

 crossed out of sight and a pencil rested between their tips, two pencils 

 will be felt in the well-known 'Aristotle's illusion'. Singleness, in the 

 realm of touch, is obviously entirely psychological in basis. In vision, 

 it is equally so — and would never have been thought to be otherwise if 

 the partial decussation had never been discovered by anatomists. 



If the total-decussators do have fusion, then as we have seen above 

 there is no reason to deny them binocular stereopsis. If there is single- 

 ness created from right- and left-eyed images, stereopsis comes along 

 with it as a sort of psychological windfall. 



The Evolution of Binocular Vision — The need for something can- 

 not operate as a cause of it; but we do have a right to ask ourselves just 

 why binocular vision has ever evolved in the first place. What does it 

 give the animal? Clearly, its adoption and extension involves a loss of 

 periscopy and must offer some compensations which outweigh that sacri- 

 fice. In ourselves, the chief advantage of binocularity appears to be a pre- 

 cision of object-localization. It does not matter that we see solidly, so 

 long as we see deeply and can say with assurance that one particular 

 billiard ball is two and one-half inches farther away than another. We 

 have this ability only because our two one-eyed images are projected to a 

 common meeting place in space; but independent convergences of our 

 two eyes would still give us parallax on an object, enabling us to locate 

 it more promptly and accurately than we can do with a succession of 

 monocular parallaxes, even if we did not perceive solidity. 



We may be sure that animals have not evolved binocularity in order 

 to see solidly. As we have seen, the percept of solidity came to them 

 as an incidental accompaniment of disparate-image-fusion. But they 

 nevertheless have had a powerful incentive to develop binocularity where- 

 ever their snouts and their beaks and their requirements of periscopy 

 would permit. This incentive was the fact that the binocular parallactic 

 cue to distance makes no demand upon intelligence. It is as automatic 

 as geometry. On the other hand, for the successful employment of the 

 monocular cues (pp. 313-4), learning to use them is a prerequisite: 



A human child must learn slowly to evaluate the size of his retinal 

 images. To him, a monster airplane a mile in the air seems like a bird a 

 few yards overhead. He has to be told why the railroad tracks seem to 

 come together, must learn the meaning of shadows. He slowly learns to 

 evaluate aerial perspective, and may be painfully deceived by it when he 



