ambulatory running round all the sides save that facing the Via Principalis At Gellygaer again, there is no sign of a wall separating the outer from the inner courtyard, but simply a single row of pillars. Even at Birrens, which probably does not date from an earlier period than the reign of Pius, a single row of pillars formed, in the original plan, the boundary between the outer and the inner courtyard. It was only at a subsequent period that the space between them was filled up with masonry. There is thus reason to believe that the type of Principia such as we find at Housesteads, with a wall and roofed passage separating an outer from an inner courtyard, is later than the type in which the outer courtyard is terminated by an arcade of pillars, and, further, that at Newstead one type was succeeded by the other.
Various Periods
The evidence we possess at Newstead does not, however, enable us to identify the Principia, of which the outlines were recovered, as belonging even in its earlier form to the first century, or, in other words, to the fort established during the period of Agricola's advance. On the contrary there is ground for supposing that the Principia of the Agricolan fort faced in the opposite directiontowards the west. Those changes in the fort which point to this conclusion will be dealt with later. In the meantime it will suffice to say that, from the remains of the building itself, it seems plain that in the alterations detailed above we have definite marks of two distinct periods, while beneath lie the traces of a still older occupation. The most definite of these traces were observed at the east end of the outer courtyard. Here, flanking the entrance, were the foundations of two walls of red sandstone (1), one foot ten inches in thickness. They consisted of broken sandstone embedded in clay, the usual river stones being absent. It was noted that the main east wall of the central building had cut through these earlier walls, and they were traced beyond it for a distance of six and a half feet under the later roadway and 'Drill Hall.' Here the two walls had been joined together by a third (1), twenty-five feet eight inches long, having a width of two feet six inches, and founded in the same way. The three walls may have enclosed the Sacellum of an earlier Principia facing the west. The suggestion that the enclosure had formed part of an older fort was confirmed by the discovery of another building, built in the same fashion and placed in alignment with it, lying beneath the adjoining buttressed building on the south. A further evidence of this older occupation was obtained in cutting through the ambulatory on the south side of the later Principia. Here, at a depth of two feet below the surface of the
52