very small Intervals of Time. 



343 



were not concordant, there appearing to be a variable and 

 rather considerable interval between the closing of the contact 

 and the motion of any electricity in the circuit. This beha- 

 viour, which had induced Sir 0. Wheatstone to lay the appa- 

 ratus aside, induced me to regard it also as not sufficiently 

 exact for the purpose for which he had suggested that I might 

 be able to use it. 



Much occupation prevented me taking up this subject again 

 until recently, when I recommenced by making a more careful 

 study of this little contact-apparatus, and found the cause of 

 its irregularity to be vibration due to the sudden stopping of 

 the impelling spring. I therefore had it reconstructed so that 

 each part was placed upon a separate base, and found that it 

 then fulfilled all that could be desired of it. 



The periphery of the rotating disk is divided into 500 equal 

 parts ; and as the finger or pointer is furnished with a nonius 

 by which ^ of a division may be accurately read, or 5 * of 

 the periphery, it follows that when the disk makes two revo- 

 lutions per second, the finger may be placed to give a blow at 

 any desired interval between 0*0001 and 0*5 second. 



The force of the spring is adjusted as nearly as can be to 

 give the requisite initial speed of rotation by observing the 

 engraved divisions of the disk, which become visible in suc- 

 cession under the fibre of a small telescope when illuminated 

 by the spark of an induction-coil, the primary circuit of which 

 is opened and closed by a half-seconds' pendulum. 



The way in which I have employed this apparatus for con- 

 trolling the above method of electrically determining very 

 small intervals of time is as follows : — The body of the disk 

 d d and spring are connected to one side of the mica accu- 

 mulator, A, fig. 3. The spring ef e, whilst trigged, rests in 

 contact with a point s, con- 

 nected with one side of the bat- 

 tery E. The remaining sides 

 of the accumulator and battery 

 are connected together. In this 

 position, therefore, the battery 

 charges the accumulator. Con- 

 tact m is joined through the 

 adjustable wire resistance r, 

 and contact n through the 

 galvanometer G, with the point 

 of junction between battery and 

 accumulator. On releasing the 

 disk the spring e, leaving s, 

 breaks the battery-circuit, and, 



