About the internal topography of the Etruscan city we know sadly little. That it was very largely determined by the natural configuration of the ground there is little room for doubt. It is true that on the Piazza d'Armi Stefani found what may have been an open square with a straight street leading out of one corner of it and a second street running for a short distance at right angles to it. But the regularity of plan extends only a very short distance back from the main façade, and it bears all the marks of being a later rationalisation of an existing irregular plan; nor is there any suggestion of a regular layout elsewhere in the city. The main lines of the street-plan are clear enough, and these indicate a radial layout, with the city-centre occupying roughly the same site as the centre of the Roman town. This was, and still is, the natural focus of the plateau. Here the crest divides into two distinct ridges, the southern one running the full length of the promontory, right down to the Piazza d'Armi, the northern one bearing off to the left and then swinging right again towards the modern Casale Cabrioli, ending on the cliffs overlooking the Fosso della Valchetta, opposite the Vacchereccia tumulus. The layout of the south-eastern part of the town was very largely determined by the course of the roads which followed these two ridges and of a third road which probably ran down the bottom of the valley between them. Two other roads, those from the Formello and the Millstream Gates, converge directly on the centre, and that from the Capena Gate joined the northern ridge-road about 500 m. to the east. The Caere road probably joined the axial road some distance to the west of the centre.