ATMOSPHERE AND THE ACOUSTIC EFFICIENCY OF FOG-SIGNAL MACHINERY. 239 
oeen said regarding the existence of echoes from fog-banks and from atmospheric 
eddies, it will be realized that this is often a matter of some difficulty. The writer 
has often observed that after a second or two from the moment that a signal is first 
heard the sound appears to reach the observer from every direction, in many cases 
before his attention can be fixed to determine its bearing. A peculiar characteristic 
of the diaphone signal which differentiates it from that of a ship siren is a short and 
powerful terminal note of diminishing pitch caused by the slowing down of the piston 
for a few oscillations at the end of a blast. This terminal note (or “ grunt,” as it is 
called at sea), being of longer wave-length, not only travels with less attenuation, 
but is easily distinguished from the reflections or echoes of the first part of the signal. 
Owing to its short duration it does not overlap its own echoes. Navigators make 
use of the first part of the signal as a “ stand-by,” and depend on the terminal 
“grunt” for fixing direction. If a signal consists of two or more blasts, the bearing 
of the fog-alarm may be determined in this way with considerable accuracy. The 
peculiarity of the diaphone note just mentioned was an unexpected development and 
is emphasized in modern installations. 
§ 13. On the Thermodynamic Measurement of Acoustic Output. 
(i.) Experimental Arrangements. 
In view of the fact that the sound-waves in the immediate neighbourhood of a fog- 
signal generator are probably propagated as waves of finite amplitude with heavy 
transmission losses, it was necessary to develop a method of measuring the actual 
proportion of power converted into sound which would be independent of the theory 
of the propagation of waves of small amplitude and of propagation according to the 
law of inverse squares. 
The principle of the method which was designed to accomplish this purpose may be 
briefly stated as follows. Imagine two thermometers inserted, one in the high- 
pressure side, the other in the side open to the atmosphere, of a siren or other sound- 
producing instrument operated by compressed air. If the air be allowed to escape 
without producing sound, there will be no fall of temperature as no external work is 
done by the escaping air (except for the negligibly small Joule-Thomson effect; the 
thermometers are supposed to be situated in regions of steady flow, free from eddy- 
currents). If, now, the siren is allowed to sound, the rate of escape of air being 
regulated to the same value as before, the low-pressure thermometer will register a 
fall of temperature, and the difference of the two thermometer readings will give a 
measure of the external work propagated to a distance as sound according to the 
theory developed in § 6. 
A reference to fig. 3 will indicate the general arrangement of the apparatus. Two 
resistance thermometers were constructed of fine silk-covered iron wire, wound inside 
a brass framework and held in position by loops of silk thread; the construction 
