When, however, the words of Frontinus have to be applied, the difficulties of the theory of his referring to a temple are increased. In the case of the expression last referred to, “_following the specus (spes)_, or, according to one manuscript, _the old specus_, which is obviously the sense,” it is very difficult to imagine what circumstances there were in connection with a temple which could warrant the use of such words; and even in the instance mentioned, where we were noticing the Appian aqueduct, “the expression, the Gemelli, which is a place under the _old specus_ (Spe̅s̅ Vetus),” is somewhat singular. Granting that a temple once existed just within the wall of the city (which, from the context, must have been its position if any) it is singular that he should use it as a landmark when describing the junction of two streams of water. The remaining three expressions are simply “at the _old specus_ (Spe̅s̅ Vetus[45]).” With these difficulties to contend with, it has been thought well to seek a different solution, and this is found in the reading of “Specum Veterem” for “Spe̅m̅ Veterem,” i.e., the “old _specus_.” With this reading, it naturally follows that it would refer to the old _specus_, or the _specus_ of the Appia and the Anio Vetus; and it is singular that the first time it occurs in Frontinus, the Codex Urbinas[46], only second in authority to the Codex Cassinensis, has the reading “Anienem Veterem,” instead of “Spem Veterem.” What was the true reading of the Codex or Codices from which these two copies were made, it is impossible to say; they are the earliest we have, and it is clear from several other instances that the scribes did not copy with much knowledge of the matter in question: then it was easy to mistake _Spc̄_ for _Spē_. That it was the Specus Vetus which was meant, must rest therefore upon the circumstances which allow of its application to the passages named, and it remains to shew that this is the case. In the instance under discussion, the water of the Anio is said to flow along its own _specus_, and therefore it would not be probably possible to find an interpretation more suitable as far as this case is concerned. The conduit of the water of the Anio Vetus had “emerged,” as the other aqueducts emerge, near the Porta S. Lorenzo, from the higher ground between the Porta Maggiore and that gate: it must consequently have been carried on a substructure from that point on the outer bank of the original fortifications of Rome, that is, on the high bank on which the aqueducts were carried, and on which the wall of Aurelian was afterwards built, to the inner bank, on lower ground faced with the tufa wall of Servius Tullius. The names of the gates are matters of dispute, and are quite immaterial; the levels of the ground decide the question[47].