DRS. GUY BARLOW AND H. B. KEENE ON THE ANALYSIS OF SOUND. 
151 
Using the full volume of the tube, the natural frequency was 40/sec., and when the 
plunger was pushed up within a few centimetres of the diaphragm the natural frequency 
became about 100/sec. There was considerable damping, as a result of which the 
natural frequency was not pronounced, and was difficult to determine. 
Listening Arrangements. —In most of the experiments it was found useful to put a 
telephone in the receiver circuit in place of the galvanometer and interrupter in order 
to listen to the sound. Such observations are recorded below as “ sound in 
telephone.” 
In the case of low-frequency vibrations, e.g. 50/sec., the sound in the telephone does 
not give a reliable indication of the character of the vibration in the water, since for 
such a frequency, the telephone responds only to the harmonics, and if they are not 
present then no sound is heard. 
A Broca tube also served as a useful listening device. The diaphragms were flat 
sheets of tinned iron, 15 cm. diameter. The natural frequency in water was approxi¬ 
mately 100/sec., and showed marked timing with depth due to change in pressure. 
The sounder 123/sec. could be heard at a distance of 400 yards. 
It was found that a short length of rubber tube, 1 cm. external diameter, walls 2 mm. 
thick, formed a simple type of non-resonant Broca tube. The lower end was sealed 
and weighted with a lead sinker, the upper end was connected to a stethoscope. 
Except the cylindrical sounder first described, both sounders and receivers were 
non-directional. 
(c) Reservoir Experiments with Sounders. 
The Barnt Green Fishing and Boating Club kindly gave permission to make use of 
Great Bittell Reservoir (near Birmingham), together with a number of boats, &c., for 
experimental purposes. The reservoir has an area of 100 acres, with a maximum depth 
of 40 feet. The receivers were connected to the shore by cable, and the analysis was 
made in the boathouse where the interrupter and 100-ohm Broca galvanometer were 
set up. With few exceptions it was found impossible even in the calmest weather 
to use the receivers when hung from a moored boat, since a very small motion at the 
point of support produced a large disturbance on the galvanometer. Satisfactory 
results were only obtained when the receiver was either hung from a tripod or placed 
directly on the bottom of the reservoir. The tripod was constructed from bamboos 
11 feet long and provided with a sinker on each foot. When in the required position, 
the lowering rope was sunk with a lead weight in order to prevent surface disturbances 
being directly communicated to the apparatus. In some experiments a small cork 
marker was used to indicate the position of the tripod. 
Determination of Natural Frequency of Sounders and Receivers in Water. —This pre¬ 
sented considerable difficulty, and it may be of interest to indicate the different methods 
which have been used. 
