the wood of a door or a chest. Fig. 22, which came from the Baths, is a strigil (Pit LVII).

Two iron objects included in Plate LXXI., Figs. 5 and 6, may be noted here. Fig. 6, found in Pit LVI, consists of two iron rings joined together by a swivel. Fig. 5, from the ditch of the early fort, is an iron collar composed of two semi-circular pieces hinged together, so that it can be opened and closed at will. The ends, which open, terminate in loops—one of these circular, the other rectangular. From examples noted on the Continent it is probable that these loops were held together by a long narrow link, to which a chain was fastened. When the collar was slipped over a wooden post or other object to which it was desired to attach it, it could be drawn taut by means of the chain. On the other hand, when the chain was slackened the long narrow link permitted the collar to be opened. These objects are to be met with on pre-Roman as well as on Roman sites. An example may be noted among the finds from La Tène,[1] while from the Limes forts they are to be seen at Zugmantel and Pfünz.[2] Precisely the same type of collar survived in the African 'slave chain.'

Nails and Bolts

Nails, bolts and fastenings were common both in the pits and as surface finds. A number of these are collected in Plate LXVII. The T-shaped objects, Figs. 1 and 4, were employed to fasten tiles and tubes to the walls. They were, therefore, common in the Baths. Most of the others tell their own story. Loops such as appear in Figs. 6 and 10 to 13, some of them with rings attached, must have been used for many purposes. Inserted in the beams of the Barrack huts, they would serve for hanging a variety of articles. The nails have a surprising air of modernity about them. They are of all sizes and shapes. Not a few are perfectly fresh. Figs. 16 to 18 represent tackets from shoes.

Weaving

Although leather was abundant in the pits, cloth seemed to have disappeared almost completely. Two small fragments were, however, discovered among the damp refuse in Pit XXIII, and it is more than probable that among the population which followed in the wake of the army and settled in the annexes, weavers were to he found. No better evidence of their presence can be adduced than the presence of the long-handled combs of bone or horn employed in weaving, which occur also at

1 Munro, The Lake Dwellings of Europe, p. 287, fig. 89, 13.

2 Der Obergermanisch-Raetische Limes, Lief. 32, 'Kastell Zugmantel,' Taf. XV. Fig. 53 ; Ibid. Lief. 14, 'Kastell Pfünz,' Taf. v. Fig. 26, and xviii. Fig. 14.

289