MEASUEEMENT OF NOISE KAYE 173 



equally loud by an audiometer test, yet the noise of one may be more 

 objectionable than that of the other. 



THE ABSOLUTE MEASUREMENT OF ACOUSTICAL ENERGY 



A standard method of measuring the absolute energy of waves in 

 general is to absorb them completely in some suitable material and 

 measure the amount of heat generated. But, even if such absorption 

 were possible in the case of sound waves, the absolute amounts of 

 energy in speech and most other sounds of everyday experience 

 are so small as to be on the border line of the capacity of the most 

 sensitive heat measuring instruments we have. For example, the 

 average speech power of the conversational voice is about 10 micro- 

 watts. This value rises to about 1,000 microwatts for the shouting 

 voice, falls to 0.1 microwatt for the quietest speech, and to about 

 0.001 microwatt for the softest whisper. To take an illustration, a 

 final cup-tie crowd of 100,000 at Wembley Stadium all talking con- 

 tinuously and rather loudly would provide as much speech-power 

 as would, if converted, light a small electric lamp throughout the 

 game. Alternatively, by the end of the match the acoustical energy 

 expended would have been sufficient, if transformed into heat, to boil 

 enough water to make one cup of tea. An especially enthusiastic 

 crowd which shouted vigorously all the time might similarly manage 

 10 cups — perhaps enough to fill the challenge cup itself ! 



There are, however, one or two outstanding examples of acoustic 

 disturbances in which substantial amounts of energy are involved. 

 Measurements in New York on ships' sirens have shown a power 

 level of about 6 microwatts per square centimeter at a distance of 

 115 feet, so that the total acoustic energy emitted by the siren would 

 appear to be about one-third horsepower. King (Phil. Trans. A., 

 1919) found an acoustic output of 1.7 horsepower in the case of a 

 fog siren. 



For sounds of ordinary magnitude, however, it is clear that the 

 outlook for thermal methods of measuring acoustic energy abso- 

 lutely is not promising, and we must turn to some other property 

 of the sound waves. The oscillatory variation of the air pressure in 

 the track of the advancing sound wave, the accompanying minute 

 changes of refractive index and of temperature, the velocity of the 

 oscillating air particles, and the radiation pressure exerted on a re- 

 flecting surface have all been employed. A conversational sound 

 corresponds to an alternating pressure (R. M. S.) of about 1 dyne 

 per square centimeter,^ in other words, to a pulsation of one mil- 

 lionth part of the atmospheric pressure. The change in refractive 



^ It is estimated that the street noise of Now York exerts an average pressure of about 

 5 dynes per square centimeter and may even reach 20. 



