34-8 Mr. H. Taylor on an Optical Illusion. . 



Theseeffects are not to be confounded with the phenomenon 

 of "accidental colours," in which the spectrum of an object is 

 of the opposite or complementary colour, these colours being 

 produced only when the eye has been exerted upon an object 

 for some considerable time; and, moreover, the complementary 

 colours of blue and red are not perceived at all in these expe- 

 riments. I mention this distinction, because an attempt has 

 been made to explain the apparent motion of these figures by 

 a reference to the theory of accidental colours. 



It remains to be shown why blue and red are the only 

 colours to exhibit this peculiarity. The only suggestion I 

 have to make is, that the intermediate tints formed by the 

 admixture of blue and red contrast strongly with both of the 

 original colours, so as to produce the required effects of light 

 and shade ; and certainly both purple and lavender form de- 

 cided contrasts with blue and red. And this is not the case 

 with any other combination of the primary colours. For ex- 

 ample, the orange tints produced by blending red and yellow 

 contrast too feebly with either of these colours for any effect 

 of light or shade, and the same objection applies to blue and 

 yellow. 



Finally, why is the illusion seen only by artificial light? 

 Not from the blue being seen as green (because some degree 

 of motion is observable in blue and red patterns by daylight 

 if they are examined in obscure corners), but from the indi- 

 stinctness of artificial light as compared with solar light ; for it 

 will be found that if the colours are held close to a lamp or 

 candle so as to be strongly illuminated, the effect is destroyed ; 

 and, on the contrary, it becomes most striking in indistinct 

 lights, as when the patterns are held under cover of the table 

 on which the lamp stands, the shadows on the red spot then 

 become exceedingly dark, while the light tints on the blue 

 spot have almost a phosphorescent appearance. For the same 

 reason, when the pattern is viewed indirectly, the effect is 

 better than when the eye is directed full upon it : and in pat- 

 terns composed of several spots or figures on their appropriate 

 ground, it is remarked that the apparent motion is greatest in 

 those parts on which the eye is not directed point blank, be- 

 cause our vision is then indistinct. 



In connexion with this subject, I may mention another phae- 

 nomenon equally curious and perplexing, to be observed with 

 these same patterns. The blue spot, when viewed indirectly, 

 always appears lighter than it really is; and when the eye 

 turns full upon it, it resumes its proper colour with a kind of 

 start : the red spot again appears much darker by indirect 

 vision, and becomes as suddenly light again when the eye is 



