Some quantitative aspects of cerebellar anatomy as a guide to speculation on cerebellar functions

Ever since it became fashionable to think of the nervous system in terms of computing machinery, the cerebellum has attracted speculators because of its limpid histology and well defined neuronal interactions (Braitenberg and Atwood 1958, Marr 1969). A reevaluation of the old timing hypothesis 25 years later (Braitenberg 1984) in the light of the wave of electrophysiology inaugurated by Eccles (Eccles et al. 1967) and of the detailed mapping of the anterior lobe performed by Oscarsson and his school will be discussed later in this paper. We want to start with more macroscopical considerations, putting in the foreground the overall shape of the cerebellar cortex as it appears with remarkable constancy, and with equally remarkable quantitative variations, in all species of mammals investigated