EXPERIMENTS WITH THE TOPLER MACHINE. 127 



unavoidable, as will appear from the following calcu- 

 lation. 



Suppose the revolving plate to have an average speed 

 of 4 1 revolutions per second, it is evident that each 

 carrier would make a complete revolution in less than 

 i second ; consequently if that were the actual duration 

 of the spark, each would be continuously visible round 

 the entire circle, and hence even a single carrier would 

 produce the bright ring. But it is only necessary to 

 this result that each should be visible until it takes the 

 place of its predecessor that is during its passage of J 

 of the circle, which reduces the time to aV of & second. 



But if they were visible even half that time, -^ of a 

 second, and each were 1J inches in diameter, and their 

 distance, from center to center, 6 inches, we would 

 have 6 ellipses, each having a length equal to twice its 

 breadth. 



From this it is evident that the smallest conceivable 

 duration of spark must produce an ellipse : but as each 

 presents the appearance of a circle, with no tendency 

 to elliptical form, the conclusion is inevitable that the 

 apparent duration of the spark is an optical illusion, 

 and that its time is so nearly zero, that it cannot be 

 estimated. 



We must conclude, then, that at the instant of dis- 

 charge the image of the carrier is photographed on the 

 retina of the eye, and at the next instant darkness 

 supervenes : but the sensation on the retina has a mo- 

 mentary duration, during which the carrier appears 

 stationary, while in reality it may have passed entirely 

 round the circle. 



It is important to notice, in this connection, that the 

 appearance and disappearance of the carriers depend 



