Another likewise very interesting ruin exists in a vault under the Capitol; to judge from the style of architecture, it cannot be of a more recent date than the age of Augustus; I have, unfortunately, not seen it myself, but a friend has sent me a description of it. Fabretti calls the tunnel of which I spoke before, an _opus priscae magnificentiae_. This is the _Fossa Cluilia_, by means of which the valley was drained; it is a work of Alban origin; its continuation towards Rome was called _Fossa Quiritium_, and is the present Marrana. From this fact we may, at the risk of not going wrong more than a hundred paces, fix the spot on which the ancients conceived the combat between the Horatii and Curiatii to have taken place: respecting this point also the most erroneous notions have prevailed. By the same means, we are enabled accurately to point out the boundary line of Latium, and the spot where the Romans thought Coriolanus to have been encamped. These facts have occurred to no one, because nobody remembered that, before Appius Claudius made the via Appia, the via Latina was the only road in that direction. The Fossa Quiritium was regarded as the work of Ancus Martius; it runs between the Aventine and Palatine into the Velabrum, and terminates in the cloacae. In most maps the walls of Rome are seen continued in the form of a triangle beyond the Tiber towards the Janiculus; the walls forming σκέλη proceeding from the Capitol and the Aventine. But this is altogether a mistake. In the age of Augustus, suburbs certainly did exist beyond the Tiber, and I have reasons for supposing that they existed there even at the period of the republic, at least in the seventh century. But it is a mistake to continue the walls so far, for the Romans had long ceased heeding the walls in extending their city. The following circumstance is a proof of this: Rome had only a single bridge across the Tiber, viz. the Pons Sublicius. Now it is said, that the Fabii went out by the Porta Carmentalis, and then proceeded across the bridge into the Etruscan territory. They passed through the Porta Carmentalis because they dwelt on the Quirinal; if they had lived on the Aventine, they would have passed through the Porta Flumentana. The bridge, therefore, evidently lay outside the walls, for otherwise they would have had to pass through two gates, and two gates would have become _nefastae_. Moreover, Varro, “De Lingua Latina,” says, that the carceres of the Circus Maximus were close to the wall of the city,[10] which in his sense is perfectly correct; for the _carceres_ cannot have been more than a stone’s throw from the wall which ran from the Capitoline to the Aventine.