line of the later Via Quintana. Such a conclusion might well have been anticipated. The forces of Urbicus about A.D. 140 would find the enlarged fort with its earthworks much as its earlier Flavian or Trajanic garrison had left it. Its reoccupation and the repair of its defences would naturally follow. Probably the opportunity would be taken to strengthen it by building the surrounding wall. It has been noted elsewhere that the overlapping ditches in front of the earlier gates on the north and south, and of the gate on the west, were filled up when the wall was built. The more substantial defence seemingly enabled the device to be done away with.

A reduction in size of the fort area followed. This might perhaps be interpreted as the result of more settled conditions on the Valium and a consequent decrease in the garrison, but it does not seem probable that the change was prompted by peaceful conditions. Rather it appears to indicate an attenuated garrison alive to the possibility of attack. The building of the reducing wall, the alteration of the main buildings, the raising of the rampart encircling the outlying Bath building were works which can only be interpreted as defensive and rendered necessary by the imminent presence of danger, a danger which ultimately caused a second, but not yet final, abandonment.

We have evidence that some eighteen or twenty years after the building of the wall, shortly before the close of the reign of Antoninus Pius, there were troublous times in Northern Britain, at the close of which forts were once more rebuilt and Rome resumed for a time the mastery.

Thus at Birrens we have distinct evidence of a reconstruction of the fort, and we have further a dedication to the Emperor by the Second Cohort of Tungrians, on a slab which no doubt formed part of a building, which gives us the date A.D. 158, and in part the name of the Governor of Britain, Julius Verus. At Birrens there is no trace of any pottery which suggests an occupation earlier than Pius. That it may yet be found is possible. But in the large collection from the site in the National Museum in Edinburgh, its absence is striking. As far therefore as our material justifies a conclusion, we must place the building of Birrens in the reign of Pius (or possibly Hadrian) and its rebuilding, shown by the alterations on its plan, in the Antonine period, and it is highly probable that this tablet of the Tungrians gives us the date. We have evidence of the same process along the Vallum, for Mr. George Macdonald's researches have proved that there also some forts underwent reconstruction in the Antonine period.[1]

1 The Roman Wall in Scotland, chaps. vii. and xii.