1897.] on Some Curiosities of Vision. 363 



the edges of the streak of light thus formed appeared to be bordered 

 with red. [Experiment shown. J 



If this experiment is performed with a strong light, the hole 

 becomes bordered with greenish-blue instead of red. With an inter- 

 mediate degree of illumination both blue and red may be seen together, 

 the blue being inside the red. 



Most of the effects that have so far been described were produced 

 by transmitted light, but reflected light will show them equally well. 

 If you place a printed book before you near a good lamp and inter- 

 pose a dark screen before your eyes, then, when the screen is suddenly 

 withdrawn, the printed letters will for a moment appear red, quickly 

 changing to black. Some practice is required before this observation 

 can be made satisfactorily, but by a simple device it is possible to 

 obliterate the image of the letters before the redness has had time to 

 disappear ; the colour then becomes quite easily perceptible. Hold 

 two screens together side by side, a black one and a white one, 

 in such a manner that there is a triangular opening left between 

 them. In the first place let the black screen cover the printing, then 

 quickly move the screens sideways so that the printed letters may be 

 for a moment exposed to view through the gap, stopping the move- 

 ment as soon as the page is covered by the white screen. During 

 the brief glimpse that will be had of the black letters while they are 

 beneath the gap, they will, if the illumination is suitable, appear to 

 be bright red. 



We may go a step further. Cut out a disc of white cardboard, 

 divide it into two equal parts by a straight line through the centre, 

 and paint one half black. At the .^«^____ 



junction of the black and white portions ^^ ^ 



cut out a gap which may conveniently 

 be of the form of a sector of about 45° 

 (see Fig. 8). Stick a long pin through 

 the centre and hold the arrangement 

 by the pointed end of the pin a few 

 inches above a printed page near a 

 good light. Make the disc spin at the 

 rate of about 5 or 6 turns a second 

 by striking the edge with the finger. 

 As before, the letters when seen through 

 the gap will appear red, and persistence 

 will render the repeated impressions 

 almost continuous. Care must be taken Fig. 8. 



that the disc does not cast a shadow 



upon the printing, and that the intensity of the illumination is properly 

 adjusted. I have here several rather more elaborate contrivances for 

 making discs rotate. 



In none of these experiments does an extended black surface ever 

 appear red, but only black dots or lines, which may of course have the 

 form of letters. And the lines must not be too thick ; if their thick- 



