The Celtic goldsmith had access to three basic sources of gold: vein or reef gold, i. e. gold mined from the ground, alluvial or placer gold panned or washed from streams.
And the recycling of gold already in circulation either within the Celtic world or from outside it, often in the form of Greek or Macedonian gold coins. Gold naturally contains a variable amount of silver, ranging from a fraction of i per cent to 30 per cent or more, the higher silver natural alloys often being referred to as electrum (Lehrberge forthcoming). During the later Bronze Age it had become standard throughout most of Europe to alloy gold with up to 10 per cent copper, perhaps to improve the mechanical properties, perhaps to counteract the whitening effect of the silver naturally occurring in the gold.
As we will see, from the eighth century BC onwards there was a change in the use of gold across much of Europe with a gradually increasing tendency to deposit it with burials. The large late bronze age hoards disappear and the gold alloys used in them were no longer available. The gold used in the Hallstatt period is unalloyed and, depending on area, could be either vein gold or placer gold (e. g. Hartmann 1987; Hofmann 1991). From the beginning of the La Tene period the copper content begins to rise again but seldom exceeds 5 per cent (e. g. Eluere 1987a; Voute 1991) until the first century BC when both coinage alloys and many of the alloys used at Snettisham must be described as ternary alloys with up to 40 per cent or more copper (Stone 1987; Northover 1992). Similar ternary alloys appear in the rather separate development of gold alloys in Iberia (Pingel forthcoming).
Around the middle of the first millennium BC the process of parting gold-silver alloys with salt to remove the silver was discovered, probably in Mesopotamia. Combined with the process of cupellation to remove base metals, this meant that gold could now be refined to better than 99.5 per cent purity (Eluere 1989a). As far as we know at present these techniques were not known to Celtic goldsmiths but high-purity gold could enter the system via the Macedonian gold coinage and some natural high purity sources (Eluere 1987b). Little thought has been given so far to the origins of silver used in the Celtic world and we await the application of lead isotope analysis more extensively. The final precious metal known to the Celts was mercury, attested by several mercury-gilded torques and other objects at Snettisham.