EXCITATION OF THE AUDITORY NERVE 499 



exactly in the mid line of the head. If it be transmitted to only one ear for a 

 time, and then the two telephones be used again, the tone appears now to be 

 on the side of the ear which was resting. If in this way the one ear be fatigued 

 for a tone say of 360 vibrations per second, and immediately afterwards one of 

 365 vibrations be transmitted to both ears, the one fatigued for 360 vibrations 

 will show no trace of fatigue for the new tone. It is difficult to see how the 

 nerve fibers could be excited directly by one of these tones and not by the other. 

 The difficulty disappears by supposing that each has a resonator which is not 

 affected by the other. 



We can think of the analysis of tones, therefore, as follows : In the internal 

 ear there are a large number of resonators adapted for different tones, which 

 are called into play if the appropriate vibrations are transmitted to the endo- 

 lymph. Each of these resonators in some way affects a nerve fiber. The 

 excitation thus aroused is transmitted to the brain and there, according to 

 the nerve fiber which brings it, gives rise to a perception of one tone or another. 



In order to test the plausibility of this hypothesis it is necessary to in- 

 quire whether the structures which might be regarded as resonators are present 

 in sufficient number to account for the analytical powers of the ear. 



Only exceptionally does one meet with a man who cannot tell definitely 

 which of two successive tones is the higher, provided that the interval between 

 them really is great enough. In musically educated individuals this ability 

 is very great. According to Preyer trained persons can recognize a difference 

 of 0.3-0.5 vibrations per second within the range from a 1 to c 11 ; above and 

 below this range the ability is much less e. g., with c y errors of as much as 

 one hundred and more vibrations may occur. 



According to Helmholtz's calculations some 4,200 resonators r. e., 600 

 per octave would be sufficient to account for the best possible discernment 

 of fractions of a half tone. Besides this, 300 resonators would be enough for 

 the tones not used in music i. e., 4,500 in all. 



We have seen that the semicircular canals and the otolith sacs probably 

 have no acoustic functions, or that at most they take part only in the percep- 

 tion of noises (cf. page 475). The whole structure of the nerve endings in 

 the cochlea, on the other hand, favors the view that the peripheral organ for 

 the analysis of sound is to be sought here. 



On the basilar membrane (Fig. 199, mb) we find the organ of Corti. This 

 contains a very large number of rodlike structures, the pillars of Corti (ic and 

 ac), standing side by side throughout the whole length of the cochlea and bound 

 together by means of a joint at the top into pairs. 



These pillars are surrounded outside and inside by peculiar epithelial cells, 

 some of which, the outer (ah) and inner (ill) hair cells, bear hairlike processes 

 ending freely in the endolymph. These cells are in connection with the end- 

 ings of the auditory nerve. The basilar membrane is of varying width at differ- 

 ent parts of the cochlea and contains fibers which are stretched transversely to 

 the cochlear canal. These are imbedded in a transparent matrix. 



The required resonators must be found among these structures and their 

 number is quite sufficient for the purpose; for, according to Retzius, the cochlea 

 of man contains 5,600 inner pillar cells, 3,850 outer pillar cells, 3,500 inner 



