Representation of Three‐Dimensional Space in the Vestibular, Oculomotor, and Visual Systems

Anatomists have long been fascinated by the arrangement of the semicircular canals, especially as there are only minor variations in size and orientation over a wide variety of species. Every skull that survives, still has the arrangement of the bony part of the labyrinths even long after all traces of its living contents have vanished. The art and technique of preserving not only the bony but also the membranous part of the labyrinths is presented in Gray's The Labyrinths of Animals (1907-1908)' in which more than 70 plates of stereoscopic photographs document the exact three-dimensional anatomy of the membranous labyrinths of various species (FIG. 1). In the first monograph about the vestibular system, Fundamentals of the Theory of Motion Sensation (1875)* Ernst Mach considered several aspects of vestibular physiology that have only recently been explored again. One is multisensory interaction, the other, how direction is coded within the sense of motion. FIGURE 2 shows his schematic drawing of the canals and their spatial arrangement. Physiology in the last century relied mostly on behavioral studies after selective lesions. In 1830 Flourens3 wrote about performing selective canal lesions in pigeons. He accurately described the instability of head movements in the plane of the lesioned canals, but gave very little thought to its interpretation. Experimental work on the labyrinths reached a climax with the techniques of Ewald (1892); who invented and skillfully performed the technique of canal plugging. As shown in FIGURE 3, the idea is to block any fluid movement or pressure build-up across the cupula during angular acceleration without interfering with the hair cells or the afferent nerve. When this is done successfully, the animals do not exhibit spontaneous or pathological nystagmus. Only after single-neuron recordings posed new questions and required new interpretations were some of the above-mentioned techniques virtually reinvented to explore pathophysiological mechanisms. Money and Scott5 reintroduced canalplugging as an experimental tool, and Simpson6 took up the art of creating labyrinthine casts (FIG. 4). Measurements of exact canal configurations and effective muscle pulling directions exist only for a few animals7 FIGURE 5 shows such measurements, using the rabbit as a lateral-eyed animal and the cat as a frontal-eyed animal.' One immediately sees that canals and muscles are approximately aligned, but that their orientations are by no means identical. It is important to note that such deviations are not just the outcome of statistical variations, but are systematic. As a result, if the vestibulo-ocular