138 DRS. GUY BARLOW AND H. B. KEENE ON THE ANALYSIS OF SOUND. 
The amplitude of the current i 0 in c.g.s. units was calculated from the formula 
JN . 27 -na 
= v/Pd + l^V'L 1 " 
where, 
J = magnetic flux cut by one turn of the coil per cm. displacement, directly 
determined by means of a ballistic galvanometer. 
N - - number of turns on the coil, usually 1-10 turns. 
n = frequency. 
a — amplitude of motion of the magnet, determined by a microscope with eye¬ 
piece scale. 
R = resistance of the circuit (9 ohms) in c.g.s. units. 
L = self-induction of the circuit. The only appreciable self-induction was due 
to the Broca galvanometer (-0073 henry). 
On interruption the current measured by the galvanometer response is i Q jn, and 
from the known sensitiveness of the galvanometer (35 div. per microampere) the magni¬ 
tude of the response can be calculated and compared with the observed value. 
The results for experiments made with frequencies ranging from 1-2000/sec. are 
given in the following table :—- 
Frequency. 
n. 
Amplitude of 
motion. 
a. 
Amplitude of 
current, 
io- 
Galvanometer 
deflection 
(observed). 
Galvanometer 
deflection 
(calculated). 
1-07 
cm. 
0-200 
amp. 
0-67 X IO -6 
70* 
73 
6-4 
0-170 
3-4 
35 
38 
11-5 
0-210 
3-1 
38 
34 
14-9 
0-200 
3-8 
47 
43 
27-5 
0-210 
7-4 
70 
83 
95 
0-200 
11 -0 
120 
120 
500 
0-025 
14-3 
190 
160 
990 
0-002 
1-26 
20 
14 
2040 
0-003 
3-8 
42 
40 
The experimental error is likely to be greatest in the case of the two highest frequencies 
owing to the small amplitudes to be measured. Moreover, the 2040 fork had to be 
sustained by bowing. The agreement between the observed and calculated values for 
all frequencies is as good as can be expected, but it only holds if account is taken of 
self-induction, since at the higher frequencies the correction for impedance is very 
large. 
* Galvanometer with 9-4 sec. period, giving 340 div. per microampere. 
