Hastings — Optical Errors of the Hitman Eye. 411 



the retinas, which can he compensated by an alteration in the 

 convergences of the axes, and thus the phenomenon is reduced 

 to one which admits of a simple stereoscopic interpretation in 

 complete accordance with the observations. This is because 

 such a shifting — whether produced by a change in the direc- 

 tion of vision or by a change of place in the object itself — 

 partially uncovers one pupil while increasing the obstructed 

 portion of the other, in short, virtually moves one pupil inwards 

 and the other outwards. The resultant effect is like that pro- 

 duced by placing a thin wedge of glass before one eye, when 

 the red appears in advance of the blue if the thicker edge of 

 the wedge is on the nasal side ; a reversal of the wedge inverts 

 the apparent relief. 



There is, however, one phenomenon which often gives a 

 determining impulse to the interpretation in accordance witli 

 the experience of Donders, namely, the relative angular dis- 

 placement of different colors upon the retina. Thus, in my 

 own case, although quite unable to recognize anything like 

 stereoscopic relief among a series of strongly colored hgures on 

 a black background when the eyes are fixed pretty steadily 

 upon them, the relief appears very striking when I walk past it, 

 or when the object as a whole is moved to and fro. So too, in 

 the experiment of the fluttering hearts described above, although 

 there is no chromatic relief under ordinary illuminations, such 

 relief is an invariable accompaniment of the fluttering when 

 produced. As has been shown above, the angular accelerations 

 and angular velocities of moving objects in the visual field 

 would vary with their color, so that colors of greater refrangi- 

 bility would appear to change their directions from the observer 

 more slowly, exactly imitating in this particular the effect of 

 greater distance. 



The illusion described would appear of rather abstract 

 scientific interest were I not convinced that the incomparable 

 French artists of the thirteenth century had recognized it and 

 employed it for the purpose of artistic expression. Indeed, it 

 was a casual inspection of the marvelous medieval windows in 

 the great cathedral at Bourges which first turned my attention 

 to the studies embodied in these papers, and which persuaded 

 me that the one essential distinction between these antique 

 windows and their unsatisfactory modern imitations lies in the 

 knowledge, possessed by the old artists, of the effect gained by 

 an ordering of their vivid colors so that the resulting chromatic 

 binocular relief should fit the composition of their pictures. 

 As far as known to me, the most beautiful surviving examples 

 of this lovely art, as well as the most convincing support for 

 the views here presented, are contained in that unapproached 



