420 



PHYSIOLOGY 



CHAP. 



objects do not depend exclusively upon the synthesis of the single, 

 elementary excitations of the retina. 



The systematic scientific study of the errors of judgment often 

 made in estimating size by visual measurement has shown that 

 even if the latter is dependent on various anatomical and physio- 

 logical conditions of the eye it is not due to them alone, but 

 must be based on the psychical process of perception. 



Optical illusions, especially of perspective and in estimation of 

 size and distance, have been known since the earliest times. But 

 it was not till the beginning of the last century that they were 



C 



FIG. 204. Optical illusion. (Hering.) 



methodically and seriously studied by Oppel, who introduced the 

 term " op tico- geometrical illusions." Volkmann, Kundt, and 

 Helmholtz made similar studies; then Zollner ano\ Poggendorff 

 (1860) drew attention to a new class, that of illusions of direc- 

 tion, and Mliller-Lyer (1889) to the so-called optical paradox. 

 These discoveries and researches were accompanied by speculations 

 as to their cause ; various theories were put forward, including 

 that which attributed the cause of the illusions to the eye-move- 

 ments (Wundt, Delboeuf, Binet), or to irradiation (Einthoven, 

 Lehmann), or to perspective (Tiery, Guye), or to psychological 



3. I. C. ct 



FIG. 205. Optical illusions. (Hering.) 



factors (Lipps, Bemessi, Schumann). At the present time optico- 

 geometrical illusions, and the conditions by which they are pro- 

 duced, are an important subject of research in all laboratories of 

 experimental psychology. 



We can only cite the most classical examples of these illusions, 

 which have no physiological explanation. 



(a) Divided distances or spaces constantly appear to us larger 

 than empty spaces (Hering). Many instances of this allusion can 

 be adduced. The space ab of Fig. 204 is really equal to the space 

 be, but owing to the dots interposed be appears much larger. The 

 square a of Fig. 205, divided by horizontal lines, appears higher 

 than it is wide ; square b, divided by vertical lines, appears, on the 



