394 BELL SYSTEM TECHNICAL JOURNAL 



have played in recent developments of acoustic instruments has al- 

 ready been indicated. To apply such amplifying means we must 

 first of all have a device to convert sound power into electrical power. 

 By far the most important instrument of this class is the microphone, 

 which is a device that translates sound into corresponding electrical 

 currents. When a medium is traversed by a sound wave it undergoes 

 periodic variations in pressure, density, temperature and particle 

 velocity. A device which translates any one of these variations into 

 corresponding electrical currents may be classified as a microphone. 



The great utility of the carbon microphone rests upon the fact that 

 it in itself functions as an amplifier, i.e., the electrical power generated 

 is greater than that absorbed from the actuating sound wave. The 

 carbon microphone, however, has not been widely used for acoustic 

 measurements, lacking the requisite stability and constancy. After 

 amplifiers became available high sensitivity was no longer so important. 

 It became possible to develop microphones in which high sensitivity 

 was a subordinate property but which were stable and constant and 

 relatively free from distortion. 



The sensitivity of a microphone as a function of the frequency can 

 usually not be easily determined from its physical constants. It must, 

 therefore, be calibrated to be useful for general acoustic measure- 

 ments. Such calibrations are commonly made in terms either of the 

 voltage generated per unit of pressure acting on the instrument, or of 

 the voltage per unit of the pressure obtaining in a plane progressive 

 sound wave before the microphone is placed in the sound field. The 

 former is referred to as a pressure and the latter as a free field calibra- 

 tion. \'ery complete discussions of the various methods of effecting 

 such calibrations have been given by L. J. Sivian ^^ and by S. Ballan- 

 tine.^^ Unless the dimensions are small compared with the wave- 

 length the microphone will diffract the sound waves and the pressure 

 on the diaphragm will not be the same as that of the undisturbed 

 sound field; for example, at normal incidence and at frequencies for 

 which the wave-length is small compared with the diameter of the 

 microphone the pressure will be doubled. The diffraction efTect 

 exhibits itself, particularly in a variation in the response-frequency 

 characteristic with angle of incidence of the sound wave, generally in 

 not an easily predetermined manner. If the form of the instrument is 

 that of a sphere it is possible to determine this variation with angle of 

 incidence theoretically. Ballantine ^^ and also Oliver ^^ have, there- 



2' Bell Sys. Tech. Jour. X, 96 (1931). 

 22 Jour. Aeons. Soc. Amer. 3, 329 (1932). 

 23P/j3'5. Rev. 32, 988 (1928). 

 2» Jour. Sci. Imt. 7, 113 (1930). 



