406 Mr. A. G. Webster [June 10, 



WEEKLY EVENING MEETING, 



Friday, June 10, 1921. 



Sir James Reid, Barb., G.C.V.O. K.O.B. M.D. LL.D., 



Vice-President, in the Chair. 



Aethur Gordon Webster, D.Sc. LL.D. Hon.M.R.L, 

 Professor of Physics, Clark University, Worcester, Mass., U.S.A. 



Absolute Measurements of Sound. 



It is now more than thirty years since it occurred to me to devise an 

 instrument that should be capable of measuring the intensity or 

 loudness of any sound at any point in space, should be self-contained 

 and portable, and should give its indications in absolute measure. 

 By this is meant that the units should be such as do not depend on 

 time, place, or the instrument, so that, though the instrument be 

 destroyed and the observer dead, if his writings were preserved 

 another instrument could be constructed from the specifications and 

 the same sound reproduced a hundred or a thousand years later. 

 The difficulty comes from the fact that the forces and amounts of 

 energy involved in connection even with very loud sounds are 

 extremely small, as may be gathered from the statement that it would 

 take approximately ten million cornets playing fortissimo to emit one 

 horse-power of sound. 



Before we can measure anything we must have a constant 

 standard. In sound we must construct a standard which emits 

 a sound of the simplest possible character, which we call a pure 

 tone ; it will be like that emitted under proper conditions by a 

 tuning-fork, which is described by saying that the graph representing 

 the change of pressure with the time shall be that simple curve 

 known as the sinusoid or curve of sines. From this connection we 

 say that the pressure is a harmonic function of the time. Unfortu- 

 nately, the pressure change is so small that at no point in a room, 

 even when a person is speaking in a loud tone, does the pressure vary 

 from the atmospheric pressure by more than a few millionths of an 

 atmosphere. Thus we require a manometer millions of times as 

 sensitive as an ordinary barometer, and, in addition, since the 

 rhythmic changes occur, not once in an hour or day, but hundreds 

 of times per second, if Ave wish the gauge to follow the rapid changes 

 accurately, we have many mechanical difficulties. 



The problem of a standard of emission has been solved by a 



