THEORY OF THE EYE. 293 



camera obscura ; the retina is the ground of the picture, 

 the crystalline replaces the glass lens.* 



This assimilation, generally adopted since Kepler's 

 time, remains open only to one difficulty ; the camera 

 obscura, like an ordinary telescope, requires to be brought 

 to a proper focus according to the distance of objects. 

 When objects are near it is indispensable to increase the 

 distance of the picture from the lens ; a contrary move- 

 ment becomes necessary as they become more distant. 

 To preserve to the images all the distinctness which is 

 desirable, without changing the position of the surface 

 which receives them, is therefore impossible : at least, 

 always supposing the curvature of the lens to remain 

 invariable ; that it cannot increase when we look at near 

 objects, or diminish for distant objects. 



* The author seems to have left this illustration incomplete. Kep- 

 ler's suggestion of the identity of the eye with the camera obscura, 

 after all, does not touch the difficulty of the inversion of the image. 

 Nor has it been considered as completely cleared up even till much 

 later times. The solution which, it is believed, is now most generally 

 assented to is this. It is a law of our constitution, dependent on some 

 physiological principle unknown, that we refer impressions on the 

 retina to objects existing, or believed to exist, in the rectilinear direc- 

 tion from which the impression comes to the retina. Consequently, 

 as rays cross at the pupil, an impression arriving at (r) in the direction 



of the arrow, will convey the idea of an object existing at (q) ; in other 

 words, a ray falling on the upper part of the retina suggests an object 

 lying below, or an inverted image suggests an erect object. 



