Some remarks on the geometry of systems