414 Prof*. 8. P. Thompson on the Function of 



(j) The theories of Mach and Lord Rayleigh agree with 

 that of Steinhauser in one respect. They assign, it is true, 

 the perception of direction of a sound to the effect of resonance 

 or of diffraction in partially resolving a compound tone ; but 

 they make the perception of this splitting-up of the sound to 

 depend upon the unequal intensity with which the resolved 

 sounds are received in the two ears. 



(£) The experiments of LordRayleigh show that our percep- 

 tion of direction is much more accurate for compound sounds 

 than for simple tones. 



(/) The experiments of the writer with the pseudophone 

 show that the acoustic illusions are produced much the most 

 satisfactorily when shrill sounds, such as the sound of a whistle 

 or the click of a metronome, are employed. 



(m) Experiments made with very low tones show that for 

 these the sense of direction is extremely inaccurate. In this 

 case the extreme effect of diffraction would cause the sound 

 to be heard almost equally loudly by both ears in all posi- 

 tions. 



10. It would therefore appear that Steinbauser's theory 

 may be true for shrill and simple tones, and therefore true 

 for the particular case of the lark. But for baser tones and 

 for compound tones it cannot be regarded as true. For 

 these the resolution-theories of Mach and Lord Bayleigh are 

 much more nearly in accord with the facts of observation; 

 though it probably yet remains to be determined whether the 

 resolution of compound tones supposed in those theories be 

 due, as Lord Rayleigh assumes, to diffraction round the 

 head, or, as Mach assumes, to the pinna and its convolutions 

 acting as resonating-cavities to reinforce certain of the partial 

 tones of the compound sound. The former is the more pro- 

 bable in the opinion of the present writer; but it is not without 

 its difficulties. 



If in binaural hearing the direction of sound were to 

 be estimated only by the process of moving the head until the 

 line of sight coincided with that of the sound, then Stein- 

 hauser's theory accounts best for the facts. 



If, however, the direction be estimated while the head is 

 held immovable, then the resolution theories are certainly the 

 more satisfactory. 



In view of the fact (as apparently shown by Graham Bell's 

 experiments) that even in monaural audition there is a fair 

 perception of the direction of sound, Steinhauser's theory 

 utterly fails to have any significance; while the theories of 

 Mach and of Lord Rayleigh are amply sufficient for the 

 facts. 



