PAPERS PRESENTED AT GENERAL SESSIONS 33 



the partitions being studied, can be measured with a 

 high degree of accuracy. 



From these measurements, the relative intensities of 

 the sound upon opposite sides of tlie partition being 

 studied are easily determined. Seated in the Sound 

 Chamber, the observer causes any desired pipe to speak. 

 A single electric contact stops the sound and starts the 

 chronometer. This contact is maintained as long as the 

 sound is audible and is broken the instant the sound 

 ceases to be heard, the time being automatically recorded. 

 This operation is repeated in the test chamber, with the 

 partition to be studied intervening. Now it has been 

 shown that although at any one point the sound fluctuates 

 through maximum and minimum values as it dies away, 

 yet on the average taken throughout the room, the in- 

 tensity decreases according to a logarithmic law, that is, 

 the average intensities at successive equal intervals bear 

 a constant ratio to each other. Hence the logarithms of 

 the ratio of the intensities on the two sides of the parti- 

 tion is proportional to the difference of the times during 

 which the sound remains audible in the Sound Chamber 

 and in the Test Chamber. The factor of proportionality 

 involves only the absorbing power of the Sound Chamber, 

 its dimensions and the velocity of sound. The first of 

 these is known from an initial calibration, made with 

 sources of sound of known acoustical outputs. Illustra- 

 tive of the general method, the computations are given 

 for a wall made of two inch solid gypsum block plastered 

 on both sides. (Figure 4.) 



Calculation of Relative Intensities of Sound, pitch 512 

 d. v., on opposite Sides of a Wall made of 2" solid 

 Gypsum Block, plastered on both sides with Gypsum 

 Plaster. 



Ii = Intensity inside 



I, = Intensity outside 



ti^time of residual sound inside = 13.48 seconds 



tj — time of residual sound outside = 8.44 seconds 



a = absorbing power of room = 4.40 



Ii 

 log,„ — = .126a (ti — t,) = .126 X 4.40 X 5.04 = 2.80 



Ii 



— = 631 



I, 



Fig. 4. 



