174 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1932 



index for such a pressure variation is about one part in a thousand 

 million, the temperature variation about one one-thousandth degree 

 centigrade, the particle velocity of the air about one-fortieth centi- 

 meter per second, and the radiation pressure only some 10"^^ of an 

 atmosphere. The corresponding power is about one one-thousandth 

 microwatt per square centimeter, it being recalled that the power 

 (or energy) varies as the square of the j)ressure and the amplitude. 



For the purposes of absolute measurement we are led to look for 

 a measuring instrument which will give readings independent of 

 both frequency and wave form of the sound. One of the most con- 

 venient is the Rayleigh disk, which measures the air or particle 

 velocity. 



The Rayleigh disk (suggested by the late Lord Rayleigh in 1882) 

 consists of a thin circular disk hanging vertically from its edge by 

 a torsion thread, the disk being of small diameter compared with 

 the wave length of the sound to be measured. Such a disk when 

 placed in a sound field experiences a couple which tends to set the 

 disk broadside on to the direction of the sound waves, much as a 

 falling leaf tends to flutter to the ground flatwise instead of edge- 

 wise. As Konig showed in 1891, the turning couple is independent 

 of the frequency of the sound concerned, and so we are enabled to 

 calculate the particle velocity in the sound wave, provided we meas- 

 ure the deflection of the disk, and the torsional constants of the 

 system. 



If the experimental conditions are such that the sound field is of 

 known distribution (for example, pure plane or spherical progres- 

 sive waves free from boundary reflections), the oscillatory pressure 

 at any point may be readily calculated from measurements of the 

 particle velocity. 



In practice disks of thin silvered glass about 1 centimeter diameter 

 are often employed, the suspending fibers being of quartz some 15 

 centimeters long and about 5 X 10'^ diameter. The disk is mounted 

 in an inclosure the walls of which are heavily lined with absorbent. 



As would be imagined, the Rayleigh disk is a fragile instrument 

 which is to be regarded rather as an ultimate standard of reference, 

 the use of which is necessarily restricted to the standardizing labora- 

 tory. More convenient and robust instruments are essential for 

 practical purposes, and recourse is usually had to electrical micro- 

 phones, preferably of a nonresonant type. These translate acoustical 

 oscillations into electrical oscillations which are amplified by a suit- 

 able valve amplifier, and so can be conveniently and accurately 

 measured, provided proper precautions are taken. 



A variety of other devices, many of them of the mechanically 

 resonated type, have also been proposed from time to time as sound 



