File:9th century gold runic fragment from an unknown object, top view (FindID 130437).jpg

English: Cast, gold fragment.The fragment is oblong, solid and was once part of a larger object. It is roughly D-shaped in cross section and tapering in height towards one end. At the higher end, the fragment appears to have been chopped. The shorter end shows some damage and it seems that part of the underside has been cut away here. The underside of the object is flat apart from a small dent and some minor ?cracks. The upper, curved, side is divided into two panels by a band running along the middle axis. It meets another band, that frames the lower edge of the object, at the narrower end.

The fragment is engraved on both sides with runes of the Anglo-Saxon runic alphabet, or futhorc. Only three letters survive on each face: these may be transliterated respectively as (dots indicate where further letters are missing): (a) ...G D E (b) T Æ A... The runes are set in relief on a recessed background, which may originally have contained niello. It is not clear what object this very small and damaged fragment came from. Although the shape is superficially reminiscent of a seventh-century sword pommel, cut in half, such pommel fittings are hollow, not solid like this, and the fragment is also very small, and stylistically later in date.

It is not possible to make much sense from such truncated inscriptions, but two tentative interpretations may be offered. (a), which is probably the end of a word, from its form is likely to be the end of a verb in the third person, past tense. This could have thus been part of a formula, 'X verb Y', as in 'X made' or 'owned Y'. (b) is probably the beginning of a word; D. Parsons, Institute for Name Studies, Nottingham University, has suggested that it might be a form of the Old English 'teah', meaning 'tie', 'fastening' or, secondarily, 'box', 'case', 'casket', 'coffer'. It is possible but unprovable that this word might be the Y element, or part of it, in the formula suggested by the [unknown] verb on the other side. If the full inscription did follow such a formula, then it follows that an X element must have preceded the verb represented by its last three letters on side (a).

This would have implications for the original scale of the object. The small letters, cut in relief against a background possibly originally inlaid with niello, are typical of ninth-century Anglo-Saxon inscriptions, such as the Æthelwulf and 'Lancashire' rings and the piece is likely to date to that period. The fact that the fragment has been both struck and chopped up, perhaps for use as bullion, might suggest that this took place in a context of Viking activity in the later ninth or tenth century; Kelvedon is on the edge of an area of Scandinavian settlement focussed on Colchester. Scientific analysis has indicated a gold content of 91-93%, silver content of 5-7% and copper content of 1-2%. The reverse has small silvery coloured inclusions visible on the surface containing the elements osmium, iridium and ruthenium, indicating that the gold came from alluvial sources. An alluvial source for gold has been identified in at least one other ninth-century Anglo-Saxon object. Dimensions: Maximum L 0.9cm; W 0.8cm; Maximum H 0.5 cm; Minimum H 0.2cm;m Weight 3.1 grams In 2001 a fragment of Anglo-Saxon gold buckle was found on the same farm. It is reported in the Treasure Annual Report 2001, page 50 and is now within the collections at Braintree Museum.