It is probable that the whole period of its history lies within the century that elapsed between the year A.D. 80 and the earlier part of the reign of Commodus. Within this period fell the two great invasions chronicled in history—that of Agricola and that of Lollius Urbicus, and, were we to argue from a superficial view of the list of coins, we might hold that during the whole of it the occupation of Newstead was well nigh unbroken. History, however, tells us of the recall of Agricola, and indicates in no uncertain way the undoing of his work, while the expedition of Lollius Urbicus was itself a reconquest, not simply a new phase in the policy of a continued occupation. Of the intervening period of withdrawal, the debris gathered from the pits and wells at Newstead provides abundant, evidence. One hundred and seven of these receptacles were cleared out. They could be divided into two distinct classes, an earlier and a later.

Save in a very few cases, their position gave no clue to their age; pits of different periods often lay side by side. But as has elsewhere been noted, the pottery fell quite naturally into two groups, and those two groups were never mingled. The early Terra Sigillata, with its arrow-point patterns and its winding scrolls of the transition period, was never mixed with the later ware, decorated with large medallions, coarser wreaths and free figures, while differences in the coarse pottery were no less strongly marked. Between the deposits of such fragments in the two groups of pits there must have elapsed a period in the course of which the sources of supply of the pottery, no less than the fashions in its decoration, had undergone a decided change. The Terra Sigillata of Pit LXXVI, bearing the stamps of Firmo, Masculus, Sabinus, is typical of the end of the first century. That of Pit LXXII, closely adjoining, turned out from the potteries of Cracuna, Ruffus, Suobnillus, is no less typical of the second century; some of it, from its similarity to pieces in the ditch of the earthen fort at the Saalburg, might date from the end of the Hadrianic period. Similarly, the dishes of Pit LIV with the stamps DAGO, OF·COTTO and OF·IVCVN are entirely distinct from those of Pit XLIX with REGINI·M, AVITVS, RVFFI·MA, and the imperfect stamp of CINNAMVS.

It would be rash to put forward the theory that all the fragments found in association in one of these pits were necessarily contemporaneous. But there can be little doubt that most of the pits were open for a comparatively short period, and certainly no example was observed in which the forms characteristic of the two groups of pottery were found at the same level.