Most Italian towns were in ancient times situated on hills, but were then not surrounded with walls any more than the Epirot towns, but localities were chosen where a hill was naturally inaccessible, or it was made inaccessible by artificial means. The hill Moriah, on which king Solomon built the temple, was originally such a hill, and it still preserves its square form amid its ruins. The ancients at most drew a wall around the base of the hill, which was either a Cyclopean or an Etruscan (i.e., a regular) wall, so as to render it inaccessible;[8] at the top of the hill there was no wall, at most a small bulwark, but in most cases even this did not exist. A sloping road (_clivus_) with two towers at the foot led up the hill, and along it ran a portico, or two walls, usually built in a zigzag. At the top there was another gate which could be closed, and which was generally flanked by two towers, so that the access might be closed both at the foot and at the top. Such was in general the character of the Latin towns, more or less perfect, and built regularly or irregularly according to the nature of the locality; and of this kind must have been the small Latin and Sabine towns out of which arose the eternal city. These places stood quite isolated, and each had its own arx, which perfectly explains Virgil’s expression _Septemque una sibi muro circumdedit arces_; these were the strong places in Rome itself, which are so often mentioned by Livy and Dionysius. Rome, therefore, had not one arx, but seven. These seven arces were then connected by means of the agger, which extended from the Colline to the Esquiline gate. In some parts of this circumference the ancient fortification remained; for example, the Quirinal (which was so high that it was necessary to make a flight of steps, which was transferred in the fourteenth century to Araceli[9]) had one very precipitous side, which required no fortification; but from it to the Capitoline a wall was built. Thence the fortification proceeded to the corner of the Aventine. This course of the ancient walls has been mistaken by all antiquarians, except a few belonging to the sixteenth century; I discovered its real course from the nature of the circumstances; I lived in the neighbourhood, and found the remains, for on the one side of the street there runs a ridge of ruins. This wall alone prevented the Tiber from overflowing the Forum, while outside the gate the inundations were very great; when, therefore, in the seventh century the wall was neglected, the Forum and the adjoining districts, as far as the porta Carmentalis, were completely inundated. In ancient Rome, this could not have happened. The Aventine is still high enough to show, that properly it required no wall, and its precipitous side towards the river may still be seen; but from that point again a wall runs towards the Caelius, for the most part behind the ditch which is now called Marrana, but anciently (Pliny) bore the name of _fossa