268 RECENT PROGRESS OF THE THEORY OF VISION. 



An objection to the Empirical Theory of Vision might be 

 found in the fact that illusions of the senses are possible ; for if 

 we have learnt the meaning of our sensations from experience, 

 they ought always to agree with experience. The explanation 

 of the possibility of illusions lies in the fact that we trans- 

 fer the notions of external objects, which would be correct 

 under normal crnditioiis, to cases in which unusual circum- 

 stances have altered the retinal pictures. What I call ' obser- 

 vation under normal conditions' implies not only that the 

 rays of light must pass in straight lines from each visible point 

 to the cornea, but also that we must use our eyes in the way 

 they should be used in order to receive the clearest and most 

 easily distinguishable images. This implies that we should 

 successively bring the images of the separate points of the out- 

 line of the objects we are looking at upon the centres of both retinse 

 (the yellow spot), ar.d also move the eyes so as to obtain the 

 surest comparison between their various positions. Whenever 

 we deviate from these conditions of normal vision, illusions are 

 the result. Such are the long recognised effects of the refrac- 

 tion or reflection of rays of light before they enter the eye. But 

 there are many other causes of mistake as to the position of the 

 objects we see defective accommodation when looking through 

 one or two small openings, improper convergence when looking 

 with one eye only, irregular position of the eyeball from ex- 

 ternal pressure or from paralysis of its muscles. Moreover, 

 illusions may come in from certain elements of sensation not 

 being accurately distinguished ; as, for instance, the degree of 

 convergence of the two eyes, of which it is difficult to form an 

 accurate judgment when the muscles which produce it become 

 fatigued. 



The simple rule for all illusions of sight is this : we always 

 believe that we see such objects as would, under conditions of 

 normal vision, produce the retinal image of which we are actually 

 conscious. If these images are such as could not be produced 

 by any normal kind of observation, we judge of them according 

 to their nearest resemblance; and in forming this judgment, we 

 more easily neglect the parts of sensation which are imperfectly 



