352 TRANSACTIONS LIVERPOOL BIOLOGICAL SOCIETY. 



of answering the question which I was unable to answer 

 on purely physical grounds. 



If the velocity remained constant in all parts it would 

 follow that the length of the region of the organ of Corti 

 set apart for stimulation by the various tones in the lowest 

 octave of audible sounds would be just half of the whole 

 length of the organ. Above this region half the remainder 

 would serve for the next octave ; and so on, each octave 

 having thus only one half as long an area as the preceding 

 one. The number of separate nerve-ends in each unit- 

 length of the organ is, however, very much the same in 

 all parts of the organ and the possibility of distinguishing 

 a difference of pitch between two tones depending pre- 

 sumably on their power of stimulating two different 

 nerve-ends, it follows that the accuracy with which we 

 can distinguish between two tones separated by only a 

 small musical interval would be enormously greater in the 

 case of low tones than in the case of high ones. And this 

 is not the case. Very minute intervals, such as two or 

 three "cents," can be distinguished more easily in some 

 parts of the tone-scale than in others, but this difference 

 is exceedingly slight. It follows therefore that approxi- 

 mately equal lengths of the organ of Corti correspond to 

 approximately equal differences of pitch — equal that is in 

 the musical sense : or in other words the waves of the 

 basilar membrane must undergo an enormous acceleration 

 in their course along the membrane, the distance (in 

 space, not time) between two successive waves being thus 

 multiplied about 100-fold. 



Near the upper limit of audibility there is a very marked 

 falling off in the power of discriminating between tones 

 differing but little in pitch, and this means that this 

 acceleration of the wave is checked before the wave has 

 reached the very end of the organ of Corti. 



