Chap. 4 1.] Why Oljetfs arejcen uprigJrf. , 397 



It has been a matter of touch doubt and difpute by 

 what meafis it happens that we fee every object in its 

 natural upright pofition, when we know it to be in- 

 verted on the object of fenfation. To this the inoft 

 fatisfactory anfwer that can be given is, that we do not 

 fee the picture which is formed "at the bottom of the 

 tye, but the object itfelf. The picture, or rather the 

 impreflion made on the retina, is the means of feeing, 

 and therefore it does not appear of material confe- 

 quence on what part of the retina the impreffion is made. 

 We in fact fee the image in the direction of that ray 

 which conveys to us the fenfation, or rather in the di- 

 rection of the axis of that pyramid, which a pencil of 

 divergent rays forms in proceeding from any point of 

 an object. Thus in Fig. 7. we fee the point, of the 

 arrow (which is indeed depicted in the lower part of 

 the eye) in the direction of the line a, A,,?that is, in its 

 proper upright poiition. On the contrary, we fee the 

 other extremity of the arrow (which is painted on the 

 fuperior part of the retina) in the direction of b, B, 

 that is at the lower end of the' object. However, there- 

 fore, the image, which is formed, may appear inverted 

 to a perfon in faceting a natural eye, as in the preced- 

 ing experiment, ftlll the eye itfelf difcerns the object in 

 its proper and natural pofition. 



As the rays of -light are emitted or reflected from a 

 vifible object in all directions, it is evident, that fome 

 of them from every part of it mud reach the eye. Thu^ 

 the object A, B, C. (Fig. 1 i.) is vifible to an 'eye in 

 "any part, where the rays A a, A , A r , A d, A. e> 

 B a y B b y B r, B d y B e, C a, C b y C c y C d y and 

 C e can come. But though rays are reflected from 

 every point of the object to every point of the circum- 

 ambient fpace, yet it is evident, that only thofe rays 

 which pafs through the pupil of the eye can affect the 

 * fenfe; 



