Mr. J. L. Tupper on an Optical Illusion. 



425 



What has to be shown is, how the position of an object can ap- 

 pear the reverse of what it is upon the lens : this he has not 

 attempted. Another singular fallacy is where he postulates G 

 too near for vision (because its rays do not converge upon the 

 retina) , and then supposes H (because at a proper seeing-distance) 

 made visible by these self-same rays. If ray A G cannot do the 

 work of vision for G because of non- convergence on the retina, 

 how can it do it for H ? His second cone is useless unless we 

 suppose a sensorium at E. 



Again, he supposes the light which comes through the hole E 

 reflected from a surface M N beyond the hole, and hence derives 

 his second cone and his means of solution. But it is not neces- 

 sary to the success of the illusion that light proceed from any 

 such surface, or froni any distance beyond the hole ; neither is 

 such hole required, any radiant point (such as a spark) being in 

 all respects sufficient. Thus the basis of solution is cut away. 



I will now explain the phenomenon by simply following the 

 rays to the retina, first alluding to normal vision, fig. 5, where, 



Fig. 5. 



the eye seeing the pin G under ordinary conditions, every point 

 in the object delivers its cone of rays (whose base is just within 

 the pupil) to the crystalline, whence, after refraction, the rays 

 pass in another cone whose apex is in the retina (only two cones, 

 depicting the point and head of the pin, are figured). Here the 

 impress of the object on the retina is not only the reverse of the 

 object's real position (or of its position on the lens), but it is 

 moreover the strongest retinal impress that the object can pro- 

 Fig. 6. 



duce, its every point being depicted by the convergence of an 

 entire cone of rays. Now (fig. 6) trace the rays as, under the 



