- The Technology of Metalwork -
been found in French hoards. They might be formed from a single, twisted square
bar at Montan, Tarn; from a square bar chiselled into a cruciform section and then
twisted, as at Soucy, Aisne; from two plain twisted bars with simple buffer terminals
at Fenouillet, Haute-Garonne, or larger numbers of twisted, square-or round-
section wires twisted together, for example at Civray, Indre-et-Loire (Eluere 1 987b).
The culmination of forming a deeply textured torque body from varied patterns
of twisted wires is seen at Snettisham, discussed below. The torques from Montan
and Civray are also part of a remarkable phenomenon which was a speciality
of Gallic goldsmiths in the second century BC. This was the cire perdu casting of
collars, terminals and even torque bodies into deeply moulded, three-dimensional
abstract vegetal forms, the sections so produced usually being hollow with thin
walls, a triumph of contemporary casting technology. Although apparently the
results of French skills, one example has been found as far away as Gajic-
Hercegmarok in former Yugoslavia, a reminder of the geographical extent of Celtic
culture (Eluere 1985).
THE PERIPHERYWe have seen how, along the Atlantic coast of Europe, the use of gold contracted
dramatically at the end of the Late Bronze Age. In southern Britain it seems to have
vanished altogether but there was a limited survival to the north and west in the form
of ribbon torques. There is some difficulty in deciding exactly which of these torques
belong to this period. There was a limited vogue for simple, loosely twisted versions
in the Middle Bronze Age and these match other contemporary gold work in
composition. Those that can be assigned to the Iron Age with some certainty on
grounds of either composition or association tend to be much more tightly twisted
from carefully shaped gold strips with a variety of terminals soldered on (Eogan
1983 b). The two groups of torques may have no ancestral relationship at all; after all,
twisting a strip of gold alloy is a simple and obvious thing to do with it; certainly
there are no associations with late bronze age objects. The key iron age associations
are Somerset, Co. Galway, with a Navan-type fibula and other obviously La Tene
bronze, and at Clonmacnois, Co. Offaly, with an elaborate, tubular, buffer-terminal
torque almost certainly made in the middle Rhine area around the beginning of the
third century BC (Raftery 1984: 175-81). These finds not only confirm the La Tene
date of the ribbon torques, but show that even areas as remote as central Ireland
had some links with important areas of the Continent. How long this peripheral
gold industry lasted is unknown. There are two or three ribbon torques of
uncertain provenance made of refined, almost pure gold, which could be dated as
late as the Roman period (Northover unpublished). There are no associations with
iron age Celtic metalwork on the mainland of Britain, and certainly none in either
the Snettisham treasure or in the Broighter hoard, which must date to the first
century Be.
The Broighter hoard is a remarkable assemblage whose find circumstances have
been the subject of controversy, although it is now accepted as a genuine hoard. It
contains undoubted imports in the form of wire bracelets from the Eastern