They also seem to have been at Bar Hill.[1] To judge from the monuments, the bow must have been comparatively short and stout. No positive traces of the weapon itself were got, but a few iron arrow-points were found, things in themselves so delicate that their survival is rare. The finest, which came from the pit in the Principia, measures from 1¾ to 2 inches long. They are triangular in section, and are slightly fluted and barbed, with a tang for insertion into the wood of the shaft.[2] Five of this type came from Pit I (Plate XXXVIII., Figs. 2 to 6), one from the barracks of the Praetentura (Fig. 7), and one from the north buttressed building (Fig. 1). The north buttressed building also yielded a larger type of arrow-head, having a point 238 inches long and a tang 138 inches long. It is square in section, 38 of an inch wide at its broadest part (Fig. 9). There is occasionally some difficulty in drawing an exact line between arrows and spears. But the small spear-like objects, measuring from 3½ to 4 inches in length and rounded at the point, probably belong to the former class. They have no harbs and are furnished with a socket for fixing them to a wooden shaft. They were evidently in use at an early period: one of them (Fig. 2), blunted by wear, came from the inner ditch of the early fort on the west side. The leaf-shaped arrow, which was present in such large numbers in the Principia at Housesteads, did not appear at Newstead. The type of short solid arrow-point shown in Plate XXXVI., Figs. 8 and 9, was not uncommon. It measured about 3¼ inches in length. The head was circular or heptagonal in section and was furnished with a socket to receive the shaft. Such weapons have been noted on many of the German Limes forts belonging to different periods, but at Newstead they were found only in the ditch of the early fort and in the early pits. More than once they were associated with the sockets terminating in a spherical projection illustrated in Plate XXXVIII., Figs. 12 and 13. Though no evidence was obtained that this association was other than fortuitous, it seems possible that the sockets were fitted to the end of the shafts. Such arrows may have served as ballista-bolts. Whether any specimen of the pilum was found is doubtful. There was certainly no representation of the long, light iron shaft of this weapon; but among the solid pointed heads there are one or two which had possibly belonged to it (Plate XXXVIII., Fig. 11).

1 The Roman Forts on the Bar Hill p. 85.

2 For the occurrence of similar arrowheads at Corbridge and elsewhere, see Archaeologia Aeliana, 3rd Series, vol. v. p. 106, and Professor Haverfield's note, Appendix II.