exception of the small quern on the extreme right, which belongs to a class entirely distinct from querns like those from Niedermendig. This came from Pit LXVI. It is made of a hard grit and belongs to the beehive type, which was rare in the fort. Professor Boyd Dawkins, dealing with Melandra Castle, has illustrated[1] a group of similar querns found there, and has attributed them to the Prehistoric Iron Age. He finds them to be identical with the querns found at Danebury, near Northampton, and at the Lake Village at Glastonbury, both dating from the period in question, but to differ from those introduced by the Romans inasmuch as the latter are thinner and wider. The Roman querns are also disc-shaped, and frequently have grooves cut in their grinding surfaces, a characteristic specially noticeable in the Niedermendig stones.

A stone mortar in very good preservation was found in Pit XCIII. Its height is six and a half inches and its diameter twelve inches. It is of the red sandstone so common on the site. On either side projects a handle two and a half inches in length. Among still smaller objects may be noted a few ballista balls from ditches and surface finds. Whetstones occurred in large numbers. They were usually made from river stones, but among them were a few which were clearly manufactured articles. The finest of these is figured in Plate LXII., and is dealt with elsewhere. One or two flat circular discs of stone should be noted. The smallest of these, one and five-eighths inches in diameter, is made of sandstone and came from Pit LXXXIX. Another is illustrated in Plate LXXXIII., Fig. 5, where there is also shown a small stone celt (Fig. 1) which was lying on the inner margin of the outer ditch system of the West Annexe. This and a flint scraper, from the ditch of the early fort, were the only prehistoric implements of stone met with in the course of the digging.

1 Melandra Castle, edited by R. S. Conway, p. 7, fig. 1.