OF ELECTRICITY AND THE DURATION OF ELECTRIC LIGHT. 589 



reduced the number of apertures to five, but with no better success, and ultimately 

 to two ; but the sound was then so feeble, compared with the accompanying noises, 

 that it could not be distinctly heard. 



The difficulty was at last overcome by employing the arm Q itself to produce the 

 sound. A small slip of paper was held to it ; and as at every revolution a blow was 

 given to the paper, its rapid recurrence gave rise to a sound the pitch of which varied 

 with the velocity of the motion. When the machinery was put in motion with the 

 maximum velocity I employed in my experiments, the sound Gj}* was obtained, indi- 

 cating 800 revolutions of the mirror in a second. I am not aware that anything can 

 have interfered with the accuracy of this result ; the same sound was heard when 

 different pieces of paper or card were used ; and on moderating the velocity, the 

 sound descended through all the degrees of the scale below it, until distinct percus- 

 sions were perceived *". 



Let us now consider what is the shortest duration of the electric light, and the 

 greatest velocity of transmission through the wire, that can be detected by means of 

 the instrument I have described. The mirror revolves 800 times in a second ; and 

 during this time the image of a stationary point would describe 1600 circles: the 

 elongation of a spark through half a degree, a quantity obviously visible, and equal 

 to one inch seen at the distance of ten feet, would therefore indicate that it exists 

 the l,152,000dth part of a second. The deviation of half a degree between the two 

 extreme sparks, the wire being, as above stated, half a mile in length, would indicate 

 a velocity of 576,000 miles in a second. This estimated velocity is on the supposition 

 that the electricity passes from one end of the wire to the other : if, however, the two 

 fluids in one theory, or the disturbances of equilibrium in the other, travel simulta- 

 neously from the two ends of the wire, the two external sparks will keep their relative 

 positions, the middle one will be alone deflected, and the velocity measured will be 

 only half that in the former case, viz. 288,000 miles in a second. 



Repeated experiments gave the following results. In all cases, when the velocity 

 of the mirror exceeded a certain limit, the three sparks were elongated into three 

 parallel lines, and the lengths became greater as the velocity of the motion was in- 

 creased. The greatest elongation observed was about 24°, indicating a duration of 

 about the 24,000dth of a second. The lines did not always commence at the same 

 places ; sometimes they appeared immediately below the eye, sometimes to the right, 

 at other times to the left, and occasionally they were out of view altogether. This 

 indetermination, it has already been explained, is owing to the arm not always taking 

 the spark at the same distance from the discharger : several discharges are therefore 

 required to be made before the eye can distinctly observe the appearances. When 



* Since this paper was read, a registering apparatus has been attached to the instrument ; it consists of an 

 index, communicating with the axis by a light train of wheels, and making one complete revolution while the 

 mirror revolves 10,000 times. The number of revolutions of the mirror indicated by this means, did not, in 

 consequence of the increased resistance to the motion, exceed 600 in a second. 

 MDCCCXXXIV. 4 G 



