have been far below the ancient surface level, a circumstance which suggests that the pit was not completely filled up until the period of the second century occupations and thus illustrates the distinction between the early and the late pottery.

As against this comparatively long list, the pits which could be confidently classed as belonging to the later periods—approximately to the middle and second half of the second century—were somewhat less numerous. The following may, however, be mentioned: Nos. I, XXIII, XXIV, XL, XLV, XLVIII, XLIX, LXX, LXXII, LXXIV, LXXX, LXXXII, LXXXIII, LXXXV, LXXXVI, LXXXVII, LXXXVIII, LXXXIX, XC, XCI, XCIII, XCV, XCVI, XCVII, XCVIII and XCIX, all of which contained the pottery associated with the same later period. Generally speaking, it may be said that late pottery was common on the surface to the south of the railway, and that most of the pits in that area probably belonged to the later periods. In none of the wells lined with wood or built with stone was there any trace of early pottery. Where they contained pottery, it belonged invariably to the later period, to which probably all of them must be assigned.[1]

Indications of Disaster

It is a curious fact that, with the exception of Pits I, XXIII and XCV, comparatively little was obtained from the later pits. Nearly every pit in the fort in which finds of importance were made can be classed with tolerable certainty as early. Was there some great disaster at the end of the first period to account for the presence in its pits and wells of so many things that can hardly have been thrown away as valueless? It is easy to understand how many worn-out objects might find their way along with the broken dishes into what were naturally receptacles for rubbish. Odds and ends of value might have now and then dropped in accidentally. But there are circumstances that rather point to deliberate concealment. Among the objects which could hardly have found their way into the pits by chance are the querns. Fragments of these were of course among the rubbish. But in each of the Pits X, XIX, XXII and LXI there was a complete quern, lying with one stone above the other and having the iron spindles still in position. All four are of the volcanic stone from Niedermendig near Andernach on the Rhine. Such things could not have been thrown away as worthless.

1 Herr Heinrich Jacobi has made the same observation with regard to the built wells at the Saalburg.