Both cerebral stereotactic surgery and cartography lack precision in monkeys when they are based on cranial landmarks as in the Horsley-Clarke system of coordinates (Percheron and Lacourly, 1973). The reliance on ventricular landmarks greatly improves stereotactic accuracy (Percheron, 1975). This is universally accepted in man. The reluctance to use ventricular landmarks in animals may be due in part to the method itself, which is strict and demanding, and to the relative complexity of the mechanical and radiological devices required (see Percheron et al., 1986). However, it is probably also linked to a misunderstanding of the theoretical bases underlying cerebral topography.