514 THUNDER-STORMS. 



ly visible as if the wheel were at rest. But if the light continue to fall upon 

 the wheel during the tenth of a second, then each spoke will continue to 

 be illuminated from the position it has the moment the light first falls upon it, 

 until it arrives at the position which the preceding spoke had at that mo- 

 ment. Each spoke will, therefore, act upon the eye while it passes through 

 the space between two successive spokes, and will, therefore, be seen at every 

 point of that space ; and as the perception it causes at any point will continue < 

 while the spoke passes through the whole of that space, it follows that the 

 wheel will appear to the eye as a flat, circular disk uniformly illuminated. 



If, however, the light continue to fall on the wheel during an interval less 

 than the tenth of a second, suppose, for example, the twentieth of a second, 

 then each spoke will be illuminated while passing through half the interval 

 between two successive spokes, and the wheel will present the appearance of 

 a circle divided into ten equal sectors, half of each sector being visible and 

 half invisible. If the duration of the light be any other part of the tenth of a 

 second, the wheel will, for the same reason, present the appearance of a circle 

 divided into ten equal sectors, a portion of each sector being visible, bearing 

 to the remaining portion, invisible, the same ratio as the duration of the 

 light bears to the difference between that duration and the tenth of a second. 

 Such an instrument will, therefore, serve as the means of estimating the du- 

 ration of any light which continues to illuminate the wheel for a period of time 

 not exceeding the tenth of a second ; and it is evident that, by varying the 

 number of spokes and the velocity of the wheel, the duration of any light may 

 be measured when its amount is greater or less than the tenth of a second. 



Such is the instrument which has been applied by its inventor to measure 

 the duration of a flash of lightning, and, also, of the electric spark. A wheel 

 consisting of a hundred spokes, dividing the space round the centre into as 

 many equal sectors, was exposed to the light of lightning during a thunder- 

 storm. By clock-work, it was made to revolve ten times per second, making, 

 therefore, one revolution in the tenth of a second, and moving through the in- 

 terval between two spokes in the thousandth part of a second. If the duration 

 of the light by which this wheel was illuminated amounted to the thousandth 

 part of a second, it would appear as a complete illuminated disk without spokes. 

 If it amounted to half a thousandth of a second, it would appear as a circle 

 divided into a hundred equal sectors, half of each sector being visible and half 

 invisible. If the duration of the light were instantaneous, it would appear as 

 a wheel with a hundred spokes stationary, in the particular position it had -it 

 the moment the light fell upon it. 



Now, such a wheel, being thus exposed to the flashes of lightning, in a storm, 

 H was found that when illuminated it always appeared stationary, though revolv- 

 iag ten times in a second. The spokes were seen distinctly, with no more 

 than their proper thickness. It, therefore, follows that the duration of the light 

 of the flashes did not amount to so great a fraction of the thousandth part of a 

 second as was capable of being appreciated by estimating the apparent width 

 of the spokes when seen by the light of the flashes. The duration of the 



flashes must then have been a very small fraction of the thousandth part of a 



( second. 



But the duration of a flash is the time which the lightning takes to move 



) through that part of space which it traverses while it is visible. Hence it fol- 



' lows, that whatever be the extent of such a distance, it is traversed in a very 



t minute 'fraction of the thousandth of a second. 



j This method of observation has only been applied to lightning of the first 

 and second kind, no opportunity having yet been found to apply it to ball-light- 

 ning 



