The Vallum and Murus always contract the width of the interval between them as they approach a river, apparently for no other purpose than a close protection of the military way, and the defence of one bridge; for if they had passed the brooks and rivers on their line at any considerable distance from each other, two bridges would have been necessary, and two sets of guards to defend them: and here it is not unimportant to remark, that the Murus always takes that brow of the ridge it traverses, which is precipitous to the north, and never deserts its straightest or most defensible course to find a convenient situation for a bridge, while the Vallum almost invariably bends inwards as it approaches a bridge, and diverges outwards as it leaves it.—_Hist. Nor._ II. iii. Horsley’s plan of the Barrier between CILURNUM and MAGNA, which is copied on Plate II., will afford several examples of the truth of these remarks. The position of the Vallum and Murus, in relation to the stations, furnishes additional evidence. The Murus usually forms the northern wall of the station, or comes up to the northern cheek of its eastern and western gates, while the Vallum protects its southern rampart, or comes up to the lower side of its doorways. The two lines give complete protection to the camps, and to the roads leading to and from them. On the supposition that the Vallum is an independent fortification, and that it was constructed nearly a century before the Wall was thought of, we must concede that its plan was such as to give the stations the least possible support, to leave them, in short, in a great measure exposed to the enemy. The manner in which the two walls combine in giving strength to a station, is very well shewn in Warburton’s plan of the works in the vicinity of CILURNUM (Plate II). It is scarcely possible to deny the justice of the remark, which he appends to the title—‘A Plan of CILURNUM ... with part of the Plan of Severus’ Wall and Hadrian’s Vallum, shewing how they are connected at the stations, and by their mutual relation to one another, _must have been one entire united defence or fortification_.' [Sidenote: SEVERUS REPAIRED THE WALL.] It is not improbable that Severus may have repaired some portions of the Wall, and perhaps added some few subsidiary defences. Richard of Cirencester gives us correct information upon several points connected with Roman Britain, which we do not learn from other authors; it is not unlikely that his view of the subject of our present study may be the correct one. He says—