The building of Severus can only be referred to the northern structure. In the first place, the structure of Hadrian was of such a nature that any sort of restoration of it could not possibly be conceived as a new building, as is said of the wall of Severus; while the structure of Pius was a mere earthen rampart (_murus cespiticius_, _Vita_, c. 5), and such an assumption in its case creates less difficulty. Secondly, the length of Severus's wall 32 miles (Victor, _Epit._ 20; the impossible number 132 is an error of our MSS. of Eutropius, viii. 19--where Paulus has preserved the correct number; which error has been then taken over by Hieronymus, _Abr._ 2221; Orosius, vii. 17, 7; and Cassiodorus on the year 207), does not suit Hadrian's wall of 80 miles; but the structure of Pius, which, according to the data of inscriptions, was about 40 miles long, may well be meant, as the terminal points of the structure of Severus on the two seas may very well have been different and situated closer. Lastly, if, according to Dio, lxxvi. 12, the Caledonians dwell to the north and the Maeates to the south of the wall which divides the island into two parts, the dwelling-places of the latter are indeed not otherwise known (comp. lxxv. 5), but cannot possibly, even according to the description which Dio gives of their district, be placed to the south of Hadrian's wall, and those of the Caledonians have extended up to the latter. Thus what is here meant is the line from Glasgow to Edinburgh.] [110: _A limite id est a vallo_ is the expression in the _Itinerarium_, p. 464.] [111: The chief proof of this lies in the disappearance of this legion, that undoubtedly took place soon after the year 108 (_C. I. L._ vii. 241), and substitution for it of the 6th Victrix. The two notices which point to this incident (Fronto, p. 217 Naber: _Hadriano imperium obtinente quantum militum a Britannis caesum?_ Vita, 5, _Britanni teneri sub Romana dicione non poterant_), as well as the allusion in Juvenal, xiv. 196: _castella Brigantum_, point to a revolt, not to an inroad.] [112: If Pius, according to Pausanias, viii. 43, 4, ἀπετέμετο τῶν ἐν Βριτταννίᾳ Βριγάντων τὴν πολλὴν, ὅτι ἐπεσβαίνειν καὶ οὗτοι σὺν ὅπλοις ἦρξαν ἐς τὴν Γενουνίαν μοῖραν (unknown; perhaps, as O. Hirschfeld suggests, the town of the Brigantes, Vinovia) ὑπηκόους Ῥωμαίων, it follows from this, not that there were Brigantes also in Caledonia, but that the Brigantes in the north of England at that time ravaged the settled land of the Britons, and therefore a part of their territory was confiscated.] [113: That he had the design of bringing the whole north under the Roman power (Dio, lxxvi. 13) is not very compatible either with the cession (_l.c._) or with the building of the wall, and is doubtless as fabulous as the Roman loss of 50,000 men without the matter even coming to a battle.] [114: The division results from Dio, lv. 23.]