186 SOUND fraction of the mean length of the sonorous waves which traverse it; for if we take 4 metres as the mean length of the waves which are propagated through the scalre, and 59 mil- limetres as the length of the united scalro, it follows that the latter is only fa of the mean wave length. Now if we imagine the seal straightened, and as forming one continuous tube with a free communication at the helico- trema, then the mean wave traversing them' will cause only fa of the lateral action which this same wave would produce if the scaloe had the length of one half of the wave ; and it fol- lows that the whole liquid of the seal will vibrate forward and backward almost as an incompressible mass, approaching in character the oscillations of a solid piston in a cylinder ; therefore, the action against the walls of the ductus cochlearis will be very slight. But now consider the change in effect on the ductus which takes place when it, together with the scala), is wound up into such an ascending spiral as really exists in the ear. The mole- cules of the liquid in the seal, thrown forward and backward by the vibrations of the stirrup bone, tend to move in straight lines, but the curved form of the scalse causes them to press against the outer or peripheral part of the upper wall (membrana Reissneri) of the duc- tus cochlearis and against the outer part of the lower wall (membrana basilaris) when the stirrup bone moves inward, and when it moves outward this action of compression is relieved from the two opposite walls of the ductus. But these actions on the walls of the ductus, produced by the vibrations of the stirrup bone, are opposed to each other, and since they take place simultaneously and with about the same intensity (by reason of the assumption of the free communication of the seal), the hair-cell chords cannot vibrate, but will only experience compressions and dilatations like the fluid in which they are immersed. Therefore, there appears a physical basis for the opinion that either there is no communication between the Bcalae, or if one exist it must be through a very constricted passage. Indeed, if we adopt the latter view, then everything works to produce the maximum effect upon the co-vibrating chords of the organ of Corti; for, when the stirrup bone moves inward, the pressure is thrown upon the outer bprder of the upper wall or roof of the ductns, thence across to the peripheral portion of the basilar membrane. This action, we may say, takes place simulta- neously throughout the whole length of the ductus, moves downward the floor of the basi- lar membrane, and thus presses the fluid of the scala tympani against the sound membrane and moves this membrane outward. But when the stirrup bone moves outward, the pressure is relieved from the elastic basilar membrane wliirh is now moved upward, while the round membrane moves inward. There are also other anatomical facts besides the inclination of the membrana Reissneri to the plane of the membrana basilaris, and the inclination of both these membranes to the plane perpendicular to the axis of the cochlea, which favors an opinion that the outer or peripheral part of the basilar membrane receives the main part yf the vibra- tions which enter the ductus cochlearis. The auditory nerve fibrils are not attached to the Corti rods or pillars, as was formerly imagined; and hence these bodies cannot be the co-vibra- Hing parts of the ductus ; but the Corti pillars appear to act as supports for the lamina reticu- laris, between which and the basilar membrane are steadily and tensely stretched the hair-cell chords, and to these chords are attached the auditory nerve fibrils. The very fact that the number of these hair-cell chords increases with the higher development of the ear, shows their important functions; for, while in man they are arranged alternately in five rows and num- ber 18,000, in other mammalia there are only two or three rows. These hair-cell chords are more perpendicular to the basilar membrane than the Corti rods, and are also different in their forms, having swellings in the middle of their lengths. These swellings must cause them to act like loaded strings, and thus each hair- cell chord is peculiarly well adapted to co-vi- brate with only one special sound. And these hair-cell chords are so directed in reference to the sound pulses which enter the ductus that their lengths are in the direction of these pulses, and therefore they cannot be directly set in motion by these vibrations. Indeed, they appear to hold the same relation to those vibrations as the antennal fibrils of the mos- quito bear to sound vibrations which exist in the directions of these fibrils. The writer has shown by direct experiment (" American Jour- nal of Science," August, 1874) that in these conditions the fibrils of the mosquito remain at rest, although when the same sound pulses fall athwart the fibril it may be set into energetic vibrations. The hair-cell chords, therefore, cannot be set into vibration by the action of the feeble pulses which may reach them direct- ly through the membrana Reissneri from the scala vestibuli ; and furthermore, the shielding influence of the membrana tectoria tends to prevent this direct action on the chords. If this view be correct, that these chords receive their vibrations from the basilar membrane, to which their ends are attached, and not directly from the impulses sent into the ductus, it ne- cessarily follows that these chords bear to the membrane to which they are stretched the same relation as stretched strings bear to the vibra- ting tuning forks to which they are stretched in directions perpendicular to the lengths of the forks. Hence it follows that a chord in the ductus will vibrate only half as often as the basilar membrane to which it is fastened. As the basilar membrane, the tympanic membrane, and the air contiguous to the latter vibrate to- gether, it follows that the auditory nerve fibrils vibrate as frequently as the tympanic mem- brane and the molecules of air outside of the
