ATMOSPHERE AND THE ACOUSTIC EFFICIENCY OF FOG-SIGNAL MACHINERY. 241 
conditions of feeble intramolecular attractions. The experimental conditions for 
realizing this result is that the two thermometers be placed in regions of flow 
sufficiently distant from the jet that all eddy-motion has disappeared and the 
equivalent kinetic energy of mean-motion has been degraded by the action viscosity 
into mean thermal molecular agitation, i . e . into heat returned to the gas. 
(ii.) Sources of Error in the Measurement of Temperatures. 
In the case of the diaphone the energy propagated away as sound must be included 
as external work done by the air under compression, and there will result a corre¬ 
sponding drop in temperature between the pair of differential thermometers. 
In the actual experiment it was not possible to place the thermometers in regions free 
from eddy-motion. Moreover, the velocity with which the air stream impinges on 
the wires of the thermometer (velocity about 42 metres/sec.) is sufficient to cause a 
slight rise of temperature, although this source of error is eliminated to a con¬ 
siderable extent by the differential arrangement. The rises of temperature due to 
this cause are included in the differential temperature measurements B and B', and are 
denoted, when the siren is sounding and silent, by SB and SB' respectively. By 
adopting the method of taking temperature readings with the siren sounding and 
silent, the source of uncertainty due to the effect of “ kinetic heating ” is reduced to 
a minimum. 
Let w +f represent the rate at which the energy of the compressed air is converted 
into mechanical effect in the diaphone when sounding. We then have, by a slight 
extension of (44), 
w+f= JC„M (B-AB) .(55) 
In this equation w is the rate at which energy is propagated away as sound in 
ergs/sec. and f includes rate at which energy is converted into vis-viva of eddy- 
motion in the region where the temperature difference 0 is observed, plus the rate at 
which energy is dissipated by thermal conduction into the diaphone piston. The 
temperature difference due to the Joule-Thomson effect is denoted by A B ; C v is the 
specific heat of air per unit mass at constant volume and M is the total rate of air 
consumption in grammes/sec. If we denote by accented letters the corresponding 
quantities referring to a measurement of temperature-differences carried out in the 
same way with the diaphone silent and adjusted so that the air consumption M 
I'emained the same at the same pressures, we have w' = 0, since there is no external 
work done as sound. Thus 
f = JC,M(ff-Aff).(56) 
From (55) and (56) we obtain 
w = JC„M[(0 —ff) —(AO —Aff) ]-(/-/'). (57) 
The extent to which the right-hand side of (57) maybe taken to represent the 
energy propagated away as sound depends on how far we may identify the term J’ 
