( 'MAI-, iv.] HKAIMNC. 1015 



since it is in the highest degree improbable that birds are able 

 to recognize tones in some manner absolutely different from 

 that employed by mammal-. 



In the face of these difficulties it has been suggested that the 

 basilar membrane, which is present in birds as well as in niam- 

 jiuils, and which, being tense radially but loose longitudinally, 

 i.e. along the spiral of the cochlea, may be considered as con- 

 sisting of a number of parallel radial strings, each capable of 

 independent vibrations, is the sought-for organ of analysis ; for 

 it may be shewn mathematically that a membrane so stretched 

 in one direction only is capable of vibrating in such a manner. 

 And the radial dimensions of the basilar membrane increasing 

 as they do upwards from the bottom of the spiral to near the 

 top give a much greater range of difference than do the rods of 

 Corti. According to this view, when a composite vibration 

 sweeps along the cochlea it throws into sympathetic vibrations 

 those small portions and those portions only of the basilar mem- 

 brane, the vibrations of which correspond to the single vibra- 

 tions of which the composite vibration is made up ; and the 

 vibrations in turn so affect -the overlying structures, that audi- 

 tory impulses are generated in particular groups of fibrils of the 

 auditory nerve. These auditory impulses reaching the brain 

 give rise to a corresponding sensation of a particular sound. 



But the dimensions of even the basilar membrane do not 

 seem wholly adequate for the purpose ; since the latest meas- 

 urements shew that in man its range is very limited. If we 

 take the whole width of the membrane, the range is from '21 mm. 

 at the base to -36 mm. at the top, though if we take the specially 

 modified part reaching from the outer feet of the rods to the 

 spiral ligament, we get a wider range, namely, from -075 mm. 

 at the base to *126 mm. at the top. On the other hand the esti- 

 mated number of radial fibres of the membrane is very large, 

 24,000 ; and even if we suppose that several fibres always vilnuu- 

 together, this would still leave some thousands of groups of 

 strings, each group acting as an analyzer. 



In the present state of our knowledge the whole matter must 

 be left as uncertain. Even if the basilar membrane acts in some 

 such way as suggested, the other structures in the auditory epithe- 

 lium present problems as yet insoluble. The true function of the 

 rods of Corti and of the reticulate membrane of which these form 

 a part, of the cells of Deiters, of the inner hair-cells as distin- 

 guished from the outer hair-cells, as well as the reason there are 

 four rows of the latter (whereby probably the effect of the vibra- 

 tions of a group of the basilar fibres is increased) and only one 

 of the former, all these are as yet merely questions which can- 

 not be answered. 



630. Even admitting that, in some way or another, sets 

 of vibrations or, to use a more general term, sets of molecular 



