RO.MA. bnnks, (ill tlipy reached the point opposite to the walls of the Janiculiim. Of this last portion only a few fragments are now visible. On the other side of the Tiber only a few traces of the ancient wall remain, which extended lower down the stream than the modern one. Not far from the river lay the Porta Poktuensis, which Ui-ban VIII. destroyed in order to build the present Porta Portese. This gate, like the Ostiensis and Praenestina, had two arches, and the same inscrip- tion as that over the Tiburtina. From this point the wall proceeded to the height of the Janicu- liim, where stood the Pokta Aurelia, so named after the Via Aurelia (vetus) which issued from it. We have already mentioned that its modern name (Porta di S. Pancrazw) was in use as early as the lime of Procopius ; yet the ancient one is found in the Anonymous of iMnsiedlen, and even in the Liber cle Mirahilibus. The walls then again descended in a NE. direction to the river, to the point opposite to that whence we commenced this description, or between the Farnese Palace and Ponte Sisto. It is singular that we do not find any gate mentioned in this portion of wall, and we can hardly conceive that there should have been no exit towards the Vatican. Yet neither Procopius (jB. G. i. 19, 23) nor the writers of the middle ages recognise any. We find, indeed, a Transtiberine gate mentioned by Spartianus {Sever. 19) as built by Septimius Severus, and named after him (Septimiana) ; but it is plain that this could not have been, originally at least, a city gate, as there were no walls at this part in the time of Severus. Becker conjectures (de Muris, p. 129, Handb. p. 214) that it was an archway belonging to some building erected by Severus, and that it was subsequently built into the wall by Aurelius or Honorius ; of the probability of which conjecture, seeing that it is never once mentioned by any author, the reader must judge. III. The C^vtitol. In attempting to describe this prominent feature in the topography of Piome, we are arrested on the threshold by a dispute respecting it which has long prevailed and still continues to prevail, and upon which, before proceeding any further, it will be ne- cessary to declare our opinion. We have before de- scribed the Capitoline hill as presenting three natural divisions, namely, two summits, one at its NE. and the other at its SW. extremity, with a depression between them, thus forming what is commonly called a saddle-back hill. Now the point in dispute is, ■which of these summits was the Capitol, and which the Arx? The unfortunate ambiguity with which these terms are used by the ancient writers, will, it is to be feared, prevent the possibility of ever ar- riving at any complete and satisfactory solution of the question. Hence the conflicting opinions ■which have prevailed upon the subject, and which have given rise to two different schools of topo- graphers, generally characterised at present as the German and the Italian school. There is, indeed, a third class of writers, who hold that both the Capitol and Arx occupied the same, or SW. summit; but this evidently absurd theory has now so few ad- herents that it will not be necessary to examine it. The most conspicuous scholars of the Gennan school are Niebuhr, and his followers Bunsen, Becker, Preller, and others; and these hold that the temple of .Jupiter Capitolinus w.is seated on the SW. summit of the hill. The Italian view, which is directly I'.OJIA. rei contrary to thi-;, was first brought into vogue by Nardini in the last century, and has since been held by most Italian scholars and topographers. It is net, however, so exclusively Italian but that it has been adopted by some distinguished German scholars among whom may be named Gottling, and Braun| the present accomplished Secretary of the Archaeo- logical Institute at Rome. Every attempt to determine this question must now rest almost exclusively on the interpretation of passages in ancient authors relating to the Capitoline hill, and the inferences to be drawn from them; and the decision must depend on the preponderance of probability on a comparison of these inferences. Hence the great importance of attending to a strict interpretation of the expressions used by the classical writers will be at once apparent; and we shall there- fore preface the following inquiry by laying down a few general rules to guide our researches. Preller, who, in an able paper published in Schneide- win's Philolofjits, vol. i., has taken a very moderate and candid view of the question, consoles himself and those who with him hold the German side, by re- marking that no passage can be produced from an ancient and trustworthy writer in which Capitolium is used as the name of the whole hill. But if the question turns on this point — and to a great extent it certainly does — such passages may be readily produced. To begin with Varro, who was both an ancient and a trustworthy writer. In a passage where he is expressly describing the hills of Bome, and which will therefore admit neither of misapprehension nor dispute, Varro says: " Septi- montium nominatum ab tot montibus, quos postea urbs muris comprehendit. E quis Cajyitoliwn dictum, quod hie, quom fundamenta foderentur aedis Jovis, caput humanum dicitur inventum. Hie mons ante Tarpeius dictus," &c. (L.Z. v. § 41, Miill.) Hero Capitolium can signify nothing but the Cajiitolino hill, just as Palatium in § 53 signifies the Palatine. In like manner Tacitus, in his description of the Romulean pomoerium before cited: " Forumque Ro- maimm et Capitolium non a Rornulo sed a Tito Tatid additum urbi credidere " (Ann. xii. 24), where it would be absurd to restrict the meaning of Capito- lium to the Capitol properly so called, for Tatius dwelt on the Arx. So Livy in his narrative of the exploit of Horatius Codes: "Si transitum a tergo reliquissent, jam plus hostium in Palatio Capi- tolio(jue, quam in Janiculo, fore" (ii. 10), where its union with Palatium shows that the hill is meant; and the same historian, in describing Romulus consecrat- ing the spolia opima to Jupiter Feretrius a couple of centuries before the Capitoline temple was founded, says, " in Capitolium escendit " (i. 10). The Greek writers use rh KatnrdiMov in the same man- ner: 'Pu>/xvos fjifv rh ITaAaTior (faTtxw — TctTioy 5e rb Ka-KnuXLov. (Dionys. ii. 50.) Hence we de- duce as a first general rule that the term Capitolium is sometimes used of the whole liill. Secondly, it may be shown that the whole hill, when characterised generally as the Roman citadel, was also called Arx: " Atque ut ita nninita arx cir- cuinjecfu arduo et quasi circumciso saxo nitcretnr, ut etiain in ilia tempestate horribili Gallic! adventns in- columis atque intacta permanscrit." (Cic. Rep. ii. 6.) " Sp. Tarpeius Romanae praeerat arci." (Li v. i. 11.) But there is no need to multiply examples on thi,s head, which is plain enough. But, thinlly, wc must observe that though the terms Capitolium and Arx are thus used generally
