312 Mr. W. Le Conte Stevens on 



go far enough. Brewster was right in considering it inade- 

 quate, and emphasizing the fact that while the foreground is 

 seen single the background must be double ; but the quali- 

 fying statement is necessary that this is perceptible only when 

 the stereoscopic displacement is considerable. Returning to 

 the comparison with sound-perceptions, two strings of different 

 material may be mistimed so as to produce beats ; but by 

 successive alterations of tension they may be made sensibl}'- 

 unisonant, with alternately higher and lower pitch, yet the 

 resultant quality in each case is perceived to be different from 

 that of the sound from either string alone. In like manner, 

 with retinal sensations, we may pass from sensible coalescence 

 with distinctions of relief in the foreground, while the back- 

 ground is double, to similar coalescence in the background, 

 leaving the foreground double. The simultaneous relaxation 

 of muscular tension produces the impression of greater distance 

 of the new points thus combined by modified fusion ; for in 

 normal binocular vision the internal rectus and ciliary muscles 

 become relaxed in withdrawing the attention from a very 

 near body and bestowing it upon one that is distant. 



4. Physiological and Physical Perspective. 

 Wheatstone and Brewster thus separately emphasized the 

 importance of phenomena that are complementary, not con- 

 tradictory. Wheatstone's theoiy of binocular perspective was 

 based mainly upon physiological considerations; and he indi- 

 cates that he regards convergence of visual lines as invariably 

 accomplished in all binocular vision, though he cites some 

 exceptional cases in which it seems not to determine the 

 apparent position of the object viewed. Brewster's theory 

 was distinctively geometric ; he considered variation of relation 

 between visual lines to be important because the successive 

 points in the field of view were to be determined by their 

 intersection; and hence he concludes that " we see distance."* 

 Subsequent investigation now proves that the localization of 

 what is viewed in the stereoscope is in no way dependent upon 

 such intersection, but that binocularly we simply estimate 

 distance as in monocular vision. The estimate is "indeed for 

 short distances far more definite, in both normal and stereo- 

 scopic vision, than when a single eye is employed, because the 

 muscles of the two eyeballs are called into associated action, 

 independently of the sensible coalescence of slightly dissimilar 

 retinal images in the binocular eye. Certain muscular con- 

 tractions are habitually associated with certain motions of the 

 body viewed, or certain relations between its different parts. 



* Brewster on the Stereoscope, p. 50. 



