154 DRS. GUY BARLOW AND H. B. KEENE ON THE ANALYSIS OF SOUND. 
driven successively at different frequencies. Since the driving current contains har¬ 
monics the forced motion of the sounder will also contain these harmonics, and if the 
frequency of one of them is near the natural frequency, then on account of resonance 
that harmonic will be very prominent, and may be much greater than the fundamental. 
But the receiver also has a selective action, and will enhance by resonance any harmonic 
which is near its natural frequency. These effects were observed in the analysis but 
will not be further described. 
The variation of amplitude with distance was determined using the two sounders near 
their natural frequencies, and the results obtained are given in the following table :— 
Depth of 
water. 
Frequency 
of sounder. 
Depth of 
sounder. 
Depth of 
receiver. 
Horizontal 
distance. 
Response. 
Index p. 
ft. 
ft. 
ft. 
ft. 
18 
44-5 
4-5 
12 
57 
150 
190 \ 
9 / 
3-1 
17 
58 
9 
On bottom. 
60 
165 
180 \ 
<5 / 
>3-5 
The high value of the index is in agreement with the results obtained with the Metal 
Diaphragm Receiver. 
A particular source of disturbance with-the Rubber Diaphragm Receiver was found 
to be due to the vibrations caused by trains passing at a distance of rather more than 
a quarter of a mile. The effects were noticed when analysing in the region 30-40/sec., 
where on occasions definite responses of over 100 div. were obtained. These disturb¬ 
ances greatly added to the difficulty of the experiments, especially when examining 
vibrations of very small amplitude. 
(d) Sound Spectrum of Evinrude Motor (Reservoir experiments). 
A description of the motor together with the general arrangements has already been 
given (p. 149). Analysis was made with both receivers. They were either placed on 
the bottom or hung from the tripod 6 feet above the bottom. In each case a spectrum 
was obtained consisting of a harmonic series, n. 2 n, 3 n, 4 n, &c., having for its funda¬ 
mental the frequency n of the motor. In general the higher harmonics have smaller 
intensities, and only in a few special cases could measurements be extended beyond 
the first ten harmonics. 
the galvanometer to give a complete oscillation, as would be expected. The exponential recovery, which 
extended over some 15 sec., was evidently due to the air leaking into the partially air-tight button. The 
whole effect completely disappeared when a small hole was drilled through the side of the button. 
It was thought that the instrument might now be less sensitive to low frequency vibrations. An experi¬ 
ment in a tank showed that for a frequency of 100/sec. the sensitiveness was reduced to about one-half. 
It would thus appear that in the reservoir experiments the wind pressure was transmitted through the 
water to the receiver. The metal diaphragm receiver did not show these effects since its button is enclosed 
in a sealed cavity. 
