1897.] on Some Curiosities of Vision. 359 



of thing happens in a more or less marked degree whenever a dark 

 object moves across a bright background, or a bright object across a 

 dark background. 



In order to see the effect distinctly by Charpentier's original 

 method, the illumination must be strong. If, however, the arrange- 

 ment is slightly varied, so that transmitted instead of reflected light 

 is made use of, comparatively feeble illumination is sufficient. A 

 very effective way is to turn a small metal disc having an open 

 sector of about 60°, in front of a sheet of ground or opal glass behind 

 which is a lamp. By an arrangement of this kind upon a larger 

 scale, the effect may easily be rendered visible to an audience. The 

 eyes should not be allowed to follow the disc in its rotation, but 

 sJaould be directed steadily upon the centre. [Experiment.] 



The acute and educated vision of Charpentier enabled him, even 

 when working with his black and white disc, to detect the existence. 



Fig. 4. Fig. 5. 



under favourable conditions, of a second, and sometimes a third dark 

 band of greatly diminished intensity, though he remarks that the 

 observation is a very difficult one. What is probably the same effect 

 can, however, be shown quite easily in a different manner. If a disc 

 with a very narrow radial slit J^ inch or J mm. wide, is caused to 

 rotate at the rate of about one turn per second in front of a bright 

 background, such as a sheet of ground glass with a lamp behind it, 

 the moving slit assumes the appearance of a fan-shaped luminous 

 patch, the brightness of which diminishes with the distance from the 

 leading edge. And if the eyes are steadily fixed upon the centre of 

 the disc, it will be noticed that this bright image is streaked with a 

 number of dark radial bands, suggestive of the ribs or sticks of the 

 fan. Near the circumference as many as four or five such dark 

 streaks can be distinguished without difficulty; towards the centre 

 they are less conspicuous, owing to the overlapping of the successive 



2 B 2 



