14 
MR. B. RUTHERFORD ON A MAGNETIC DETECTOR OF 
metal arm LM, which pressed against the circumference of the circle and could be 
moved round it. The “detector” consisted of about thirty very fine steel wdres, 
■003 inch in diameter, arranged into a com23ound magnet about 1 ceutim. long. The 
wdi'es were insulated from each other by shellac varnish, and the small needle was 
fixed inside a thin glass tube wdiich could be easily slijDjDed in and out of the central 
glass tube OM. 
Fig. 4. 
A divided scale was jDlaced round the circumference ABC, and the vdrole arrange¬ 
ment was fixed in ^oosition before a small mirror magnetometer. The magnetized 
“ detector ” needle was placed in jDositiou by sliding it in the glass tube, and the 
deflection due to the needle was compensated by another magnet. The wires of the 
discharge circuit were connected to C and M, and when the arm ML rvas at C no 
eflect was j^roduced on the needle. When the discharge 2 )assed round the circle 
tliere was a deflection due to the partial demagnetization of the detector. The 
detector was then quickly removed and magnetized to saturation in an adjacent 
solenoid and then rej^laced. It was found that, j^rovided the detector was magnetized 
in a very strong field, on rejflacing it in position the zero remained unchanged, and 
the same deflection was obtained time after time for similar discharges. 
Since the magnetic field at the centre of a circle due to an arc of length I is 
given by 
H = 
/7 
where y is the current. 
we see that the magnetic force acting on the needle is ^proportional to the length of 
the arc traversed by the discharge. 
A series of observations were made and it was found that the deflection due to the 
detector was ajpiproximately ^proportional to the magnetic force acting on the needle, 
provided the magnetic force was well below the value required to completely 
demagnetize the steel. 
Curve (2) represents the relation between the deflection of the magnetometer and 
the magnetic force acting on the needle. The curve is nearly a straight line exceqpt 
near the tojp j^art of the curve. 
To determine the damping of the oscillations a discharge w'as passed in one direc¬ 
tion and the deflection noted. The detector was removed, magnetized and rejplaced. 
The direction of the discharge was reversed and the arnp of the circle moved until 
