produced hy varying Magnetic Induction, 473 



get up a microscopic swing in the arm or to check an 

 ah'eady existing one. 



Under these circumstances aresultmay very well be observed; 

 but it will be certainly due to a spurious cause unless far more 

 elaborate precautions are taken. 



For a whole day, List Christmas twelvemonth, I got the 

 most beautifully consistent results. The deflexion reversed 

 with the current, with the sign of the static charge, and every- 

 thing just as it ought. But before jumping to any conclusion 

 the precaution was taken of going through the operations 

 with the key while one end of the wire was disconnected and 

 the battery circuit therefore not complete. The effect was 

 much stronger. 



The whole thing was a mere electrostatic effect due to the 

 action of the slope of potential in the conducting-wire on the 

 very sensitive electrometer-like needle. 



To get over this effect was very troublesome and seemed 

 impossible. A number of obvious things were tried, such as 

 cutting the wire in the middle and adjusting its two halves 

 symmetrically ; putting the middle of the wire to earth, &c. 

 Putting a carefully selected portion of the wire to earth 

 did, indeed, get over the difficulty ; but so sensitive was the 

 arrangement that a variation of an inch in the position of the 

 earth-contact on the length of the wire made a perceptible 

 difference. The effect seemed altogether larger than what 

 we wanted to observe ; so, at length, I had the whole glass 

 box containing the charged needle coated with tinfoil. 



The result was a beautiful steadiness. The slope of potential 

 was utterly cut off and not a trace of oscillation could be 

 detected on open circuit. 



But then neither could anything be detected when the 

 circuit was closed. If the needle were at rest, it remained at 

 rest; and if it were in motion, its motion subsided without the 

 least reference to what one happened to be doing with the key. 



1 came sorrowfully to the conclusion that in cutting off the 

 electrostatic disturbance we had neutralized the thing w^e 

 were looking for. 



But on writing this view to Mr. Chattock, I found he did 

 not agree with it. His view was that the magnetic lines 

 had to get through the space in the centre of the coil, and 

 that, though a sheet of tinfoil might delay them momentarily, 

 it could make no difference in the end; since the time required 

 for the reversal of the split-up mass of iron was far less than 

 that of a semi-oscillation of the needle. 



Not feeling absolutely sure either way, I determined, at 



