- Chapter Sixteen -
plating. Choisy-au-Bac also produced ingot copper, and a touchstone and other
traces of goldworking. These two sites are only one category and others, from hill-
forts to small villages, may also have been the home of bronzesmiths. A British
equivalent of the two river-confluence sites may be Weybridge, where the Thames
and the Wey meet (Hanworth and Tomalin 1977). The metalworking evidence is
scanty but there was some very early ironworking there and an exotic, imported
ribbed bucket was found nearby. There is not too much evidence of metalworking
sites but debris from Dinorben shows that bronzeworking was a feature there as
early as this (Guilbert pers. comm.).
During the La Tene Iron Age in Britain the remains of metalworking become
much more abundant and there is a very distinct pattern to the location of different
types of process. Analysis L of the metal waste from three important hill-forts
- Danebury (Northover 199IC), Maiden Castle (Northover 1991a) and South
Cadbury (Northover unpublished) - showed it to be dominated by fragments of
sheet, either off-cuts, scrap, carefully folded packets or even partially fused masses.
Much of the debris at Maiden Castle was identified as coming from a workshop
involved with the repair or demolition, or even manufacture of typical La Tene
bronze and iron cauldrons. South Cadbury (Spratling 1970) has also been associated
with sheet metalworking and the hill-fort at Bredon Hill (Hereford and Worcester)
yielded iron hammers suitable for sheet metalworking. The metalworking evidence
from Danebury is much more limited but there was a remarkable find of what
had been a small cloth or leather bag filled with swarf from a number of copper
alloys, either from drilling or from engraving (Northover 199Ic). Crucibles or
crucible fragments have been found at all these sites but there appears to have been
no significant casting industry.
The classic La Tene period foundry site in England is at Gussage All Saints, Dorset
(Foster 1980). The excavators uncovered a pit filled with the debris of moulds and
crucibles used for the eire perdu casting of numerous items of horse harness. Other
similar sites have been found, for example at Beckford (Hereford and Worcester) and
Grimsby (Lincolnshire), both in course of preparation for publication. All of these
sites are simple settlements, at most with a simple ditched enclosure surrounding
them. None are hill-forts although there are such sites in the neighbourhood of
Gussage All Saints and Beckford. Clearly casting operations of this type were not
part of the economy of hill-forts. Village-scale industry of this type persisted until
the Roman conquest at least, as seen at a small iron age site at Stanton Harcourt
near Oxford (Oxford Archaeological Unit, unpublished). Other types of site were
involved in bronzeworking: on the Continent the great oppida such as Manching
and Mont Beuvray were certainly most important but such semi-urban sites came late
to Britain. Where they have been explored, as at Bagendon (Clifford 1961) and
Silchester, a varied metal economy is evident with signs of the working of all metals,
including the use of precious metals for the coinage. A new feature at Silchester,
Colchester and Hengistbury Head is the use of matte, otherwise copper sulphide,
a semi-smelted product brought to the site for further processing. These sites are far
removed from copper mines and the implications of the matte are still being worked
out (Salter and Northover 1992). Other special sites were also important. The
published data show a wide range of metallurgical activity at Glastonbury; at the