318
Nonę of the smali chains among the recent finds can unequivocally be identified as a dress ac-cessory, though the first one described here probably was, and all of those listed below could have been. Necklaces are not common in mediev-al art until the end of the period considered here, and bracelets at this datę are so rare that any example would be an exceptional object.
To judge from late-14th and 15th-century rep-resentations, elaborate collars for the neck were fashionable among the upper classes, often serv-ing to indicate political allegiance. The use of precious metal and highly decorative linking might suggest that a chain was intended to be wom, but there seem to be no definite criteria. Most of the excavated chains are of relatively simple manufac-ture, with plain, round or S-links. Most are of copper alloy or iron, but one is of silver. It is assumed that nonę of the plainer base-metal chains would have been wom around the neck as a piece of jewellery in its own right. Some of them could have been used to hołd groups of keys (like a modem key ring - a 16th-century group of seven keys on just such a copper-alloy chain was found at Queenhithe in London - private collec-tion) or other valued possessions - see fig 210. Eąually, this kind of chain may have been used for hanging scalę pans from a balance arm, for cens-ers, or for other purposes.
1590 SWA81 acc. no. 2397 (context 2115) ceramic phase 12 fig 210
Two thin, plain rings, linked together, each d 22mm; combined weight 1.80 grammes; one link is broken and distorted.
Perhaps part of a collar or other item of jewellery, despite the thinness of the links.
Cf AR Goodall 1983, 231 & 233 fig 1 no. 4 for a remarkably precise parallel (including the break in one link) dated to C.1270-C.1400. See Scott 1986, 83 pl 83, for a chain collar from the 1440s possibly of an appropriate type, wom by an officer of the royal household, though here the links are both heavier, and decorated. A precious-metal chain might be used to secure an elaborate badge, as with the Dunstable Swan jewel (Cherry 1987, 487-88 no. 659; see also idem 1973, 312, for discussion of chains of gold in the Fishpool Hoard), but the thinness of the two present links seems as incompatible with this as with the other functions suggested above.
1591 SM75 20 (146) 6-7 fig 210 1 107mm; each link, which consists of a ring of triply-spiralled drawn wire, interconnects only with the two adjoining ones; the wire has been obliquely cut at each end, and there are minor differences in the degree of overlap at the ends of the spirals; in most links the stack of spirals is slightly inclined, perhaps distorted from a straight stack arrangement.
Cf Mitchiner (1986, 251 no. 983) for an apparently complete chain 260mm long, joined to make a circle (found at Greenwich). This seems too long to be a bracelet as claimed by Mitchiner, but too short for a collar or to be wom doubled on the wrist. The sealed deposit in which the chain from the St Magnus site was found provides a dated context within an excavated seąuence, but the early 16th-century datę attributed by Mitchiner to this item is broadly consistent with other finds from Greenwich. Cf also Baart et al. 1977, 206, 208 & 210, nos. 379 & 385 for chains of this kind attributed to the early 17th century. This simple form of link seems to have continued to be madę for at least four centuries.
1592 TL74 2255 (1590) 9 fig 210
Four lengths (not continuous); to tal length 140mm+;
copper (AML); chain consists of eight S-links of D-