DES. GUY BARLOW AND H. B. KEENE ON THE ANALYSIS OF SOUND. 
153 
remarkable rapidity. Putting amplitude oc ——-— , the value of the index p (see 
below) is generally 3 to 3-5, which makes the energy fall off with the sixth or seventh 
power of the distance. This gives rise to considerable difficulty in making exact measure¬ 
ments, as a large error is introduced by a small change in distance. 
The variation in amplitude with distance and depth is shown in the following 
examples 
Depth of 
Frequency 
Depth of 
Depth of 
Horizontal 
Response. 
Index f. 
water. 
of sounder. 
sounder. 
receiver. 
distance. 
ft. 
ft. 
ft. 
ft. 
12 
105 
6 
On bottom. 
4 
Off scale > 250 
45 
18 
75 
3 
/ 
3*5 
15 
124 
10 
9 
40 
270 
\ 
9 
80 
25 
/ 
O ’ O 
9 
130 
<5 
\ 
/ 
>2-6 
On bottom 
80 
3 
6 
80 
16 
17 
66-5* 
3 
11 
135 
7 
\ 
3-0 
11 
60 
80 
/ 
On bottom 
60 
3 
9 
60 
150 
3 
60 
80 
Rubber Diaphragm Receiver .—When hung from a boat in very calm weather the 
galvanometer showed oscillations of 20-40 div. with an occasional 200 divs., but on 
the bottom the disturbance was reduced to 1-2 div. A residual disturbance of this 
magnitude was always present under the cpiietest conditions, even when out of the 
water, and this may represent the natural limit of steadiness of the microphone. The 
instrument was used either on the bottom or on the tripod, and even then quiet weather 
was essential. On a windy day there were large disturbances always closely associated 
with the gusts. Their magnitude was not changed by altering the natural frequency 
of the receiver. | 
With this apparatus a number of experiments were made in which the sounder was 
* In this case the weak octave 133/sec. present in the sounder was reinforced by resonance with the 
receiver. This response varied with depth in the same way as the fundamental. 
j Some experiments made in the laboratory at a later date furnished the explanation of these disturb¬ 
ances. It will be seen from the construction of the apparatus that slow pressure changes can be com¬ 
municated to the air cavity through the compensating reservoirs. Such pressure changes would not produce 
any motion of the diaphragm. The effects on the microphone are due to the direct action of the air pressure 
on the lid of the button. This was confirmed by blowing air into the cavity of the instrument. It was 
then found that when the microphone was subjected to a sustained additional pressure the microphone 
current showed a rapid increase followed by a slow exponential recovery, the transformed current causing 
