CHAP, in.] SIGHT. 161 



known. Thus when an object gives rise to a retinal image of a 

 certain size, that is to say has a certain apparent size, we estimate 

 the distance from us of the object giving rise to the image, and 

 upon that come to a conclusion as to its real size. Conversely, 

 when we see an object, of whose real size we are otherwise aware, 

 or are led to think we are aware, our judgment of its distance is 

 influenced by its apparent size. Thus when part of our field: oi 

 vision is occupied by the image of a man, knowing otherwise the 

 ordinary size of a man, we infer, if the image be very small, that 

 the man is far off. The reason of the image being small may 

 be because the man is far off, in which case our judgment is 

 correct ; it may be, however, because the image has been lessened 

 by artificial dioptric means, as when the man is looked at through 

 an inverted telescope, in which case our judgment becomes an 

 illusion. So also a picture on a magic lantern screen when 

 gradually enlarged seems to come forward, when gradually di- 

 minished seems to recede. In these cases the influence which 

 the absence of any muscular sense of binocular adjustment or 

 monocular accommodation ought to bring to bear on our judg- 

 ment, is thwarted by the more direct influence of the association 

 between size and distance. An instructive illusion of a similar 

 kind is produced by developing in the eye a strong negative 

 image ( 782) and projecting the image on to a screen which is 

 made to move backwards and forwards, or is alternately inclined 

 at various angles ; the negative image appears to change in size 

 and shape, although it is absolutely subjective in nature and 

 wholly independent of the movements of the screen. 



The complex reaction on each other of judgments as to distance 

 and size is illustrated by the experience that an object such as a 

 person looks unnaturally large when seen in a fog ; being seen 

 indistinctly, he is judged to be farther off than he really is, and 

 so appears larger than he naturally would do at the distance at 

 which he is supposed to be ; and we are similarly influenced by 

 the greater or less brightness or saturation of colours. Conversely, 

 distant mountains when seen distinctly in a clear atmosphere 

 appear small, because on account of their distinctness they are 

 judged to be nearer than they really are. The indistinctness of 

 the image of the moon or sun when seen on the horizon, similarly 

 contributes to its appearing larger than when seen in the zenith ; 

 our judgment however is probably in this case also due to our 

 being better able to compare the moon or sun with terrestrial 

 objects. We seem moreover in this matter to be especially in- 

 fluenced by our conception (which is itself an illustration of the 

 subject we have in hand) that the vault of the heavens is flatter 

 than it really is ; the zenith appears to be less distant than the 

 horizon; a geometric construction will shew that a body of the 

 same size placed at different parts of the real (spherical) vault 

 will appear greater near the horizon than near the zenith of 



11 



