RESUMO
The scanning electron microscope (SEM) was used to investigate the morphology of the neuroepithelial regions of the vestibular ampullary structures in 47 White King pigeons. The specific neural surfaces studied were (1) the cristae ampullares of the vertical and lateral membranous ampullae, (2) the hair cells lining the cristae, (3) the ampullary nerve fibers, and (4) the bipolar cells of the vestibular (Scarpa's) ganglion. Additionally, some observations of the gross anatomical structures of the bony labyrinth are given. Arguments are advanced which show that if the surface area of a given semicircular canal can be projected onto one of the three normal head planes, then that canal can be made to respond to motion in the appropriate plane, provided that the projected area is sufficiently large to achieve a threshold pressure as determined by a generalized form of Groen's equation ('57). With regard to the cristae ampullares, it is hypothesized that their surface areas can be described by means of a revolved catenary, i.e., a catenoid of revolution. (The catenary is found in nature as the approximate shape taken by a flexible cable when it is suspended at two points). The surface area of a catenoid provides a minimum surface of revolution. In the context of a crista, this implies that the given number of hair cells could not be fitted onto a smaller surface area. One advantage of this is that nature is able to utilize a thinner cupula than would be possible with other configurations and therefore an increased sensitivity to cupular motion can be realized. A second important factor is that all hair cells must revolve (by way of cupular motion) about the same centre of rotation in response to angular acceleration. Thus, all of the orthogonally-positioned hair cell tufts on the cristae surface may be stimulated simultaneously by way of a tangential shear. Other arguments show that the classical "swing door" type of cupular motion is not consistent with SEM and other recent observations. Two alternate modes of cupular motion are presented, each of which requires far less energy expenditure than does the "swing door" cupula. The suggestion is then made that, during normal head movements, the cupula behaves as a drum much like the tympanic membrane and that only for large, non-physiological motions does the "swinging door" mode of cupular motion take place. It must be remembered, however, that cupular motions during normal physiological head movements are infinitesimally small (Oman and Young, '72).