THE SENSES 



925 



are formed on identical points ; and by rapid movements the eyes 

 fix successively different parts of the field of view. (2) Vision grows 

 less distinct as we pass out from the centre of the retina, and we are 

 accustomed to neglect the blurred peripheral images in comparison 

 with those formed on the fovea. (3) When the images of an object 

 do not fall on identical points, one of the points on which they do fall 

 may be occupied with the images of other objects, some of which 

 may be so boldly marked as to enter into conflict with the extra 

 image and to suppress it. (4) Lastly, the physiological ' identical 

 point ' is not a geometrical point, but an area which increases in size 

 in the more peripheral zones of the retina, and can also be increased 

 by practice ; and images which lie 

 wholly or in chief part within two 

 corresponding areas practically co- 

 incide. 



Stereoscopic Vision. Although the 

 retinal image is a projection of ex- 

 ternal objects on a surface, we per- 

 ceive not only the length and breadth, 

 but also the depth or solidity of the 

 things we look at. When we look 

 directly at the front of a building, 

 the impression as to its form is the 

 same whether one or both eyes be 

 used, although with a single eye its 

 distance cannot be judged so accur- 

 ately. But when we view the build- 

 ing from such a position that one of 

 the corners is visible, we obtain a 

 more correct impression of its depth 

 with the two eyes. This is partly 

 due to the fact that to fix points at 

 different distances from the eyes the 

 visual lines must be made to converge 

 more or less, and of the amount of 

 this convergence we are conscious 

 through the contraction of the mus- 

 cles which regulate it. But there is 

 another element involved. When the 

 two eyes look at a uniformly-coloured 

 plane surface, the retinal image is 

 precisely the same in both. But 

 when the two eyes are directed to 

 a solid object (say a book lying on a 

 table) the picture formed on the left 

 retina differs slightly from that formed on the right, for the left eye 

 sees more of the left side of the book, and the right eye more of the 

 right side. 



That there is a close connection between uniformity of retinal 

 images and impression of a plane surface on the one hand, and 

 difference of retinal images and impression of solidity on the other, 

 is proved by the facts of stereoscopy. It is evident that if an exact 

 picture of the solid object as it is seen by each eye can be thrown 

 on the retina, the impression produced will be the same, whether 

 these images are really formed by the object or not. Now, two such 

 pictures can be produced with a near approach to accuracy by 



FIG. 403. BREWSTER'S STEREO- 

 SCOPE. 



p and TT are prisms, with their re- 

 fracting angles turned towards 

 each other. The prisms refract 

 the rays coming from the points 

 c, 7 of the pictures ab and a,(3 so 

 that they appear to come from a 

 single point q. Similarly, the 

 points a and a appear to be situ- 

 ated at /, and the points b and /3 

 at <t>. 



