no SCIENCE PROGRESS 



The intensity is determined by the factor ^-^. If with 



increasing volume and therefore increasing S the intensity is 

 not to sink below a given value a must be made small. In 

 practice a must be assumed greater than the lowest attainable 

 value since larger rooms serve to accommodate larger audiences 

 who raise the average value of a. Since then a cannot be 

 made arbitrarily small S should not exceed a certain value. 

 So the improvement of acoustics by diversification of the surface 

 has its limit. If carried too far the intensity of the sound will 

 be influenced too much. 



Since also a short period of reverberation is desirable, for 



speech about 1-5 seconds is most satisfactory, —^ must be 



4V 



chosen correspondingly large. The considerations already put 



forward show that the numerator has an upper limit. Hence 



V is similarly limited in value. Further, if a room be too large 



it is intrinsically impossible to fashion it satisfactorily for 



sounds of limited strengths, e.g. chamber music. Hence an 



orchestra hall which is quite satisfactory for orchestral music 



where A is large may be quite unsuitable for chamber music. 



With rooms for speaking where the sounds are still more 



limited in strength and where shortness of reverberation is 



even more important, it is still more necessary that they be 



not too large, and Jager considers that the poor acoustics of the 



Austrian Chamber of Deputies is predetermined in part by its 



large volume. 



Jager 's theory also shows that the maximum intensity is 



not reached until a short time after the emission of the sound, 



a time which varies with the auditorium. The law of growth 



of intensity is found to be given by the equation 



E, 



{i-e 4v'). 



— atS 



Thus E attains the value E^ when e 4V ' becomes negligible 

 compared with i. The attainment of the maximum must not 



take too long, hence —^^ must not sink below a certain value, 



4V 



the same condition as for short reverberation. In larger rooms 

 a longer time will be taken to reach the final condition, and 

 slow speaking will be advantageous. Similarly in large concert 

 halls runs and trills should be taken more slowly than in small 

 halls. 



Jager's work also indicated that in the case of a porous wall 

 there would be more absorption for high than for low pitches, 

 and Sabine extended his experiments to cover sounds varying 



