Celtic openwork vessel

Celtic gold bowl

Item description

  • Object: Openwork gold vessel
  • Date: Late 5th – early 4th century BC
  • Culture: Celtic
  • Dimensions: Height 11.5 cm; diameter across handles 17.5 cm

The vessel has an ovoid profile, with a central horizontal pierced band and two extending handles. The form and construction of the vessel, including the rolled-over top lip, show that the gold once enclosed a wooden bowl, held in place by the same gold nails that fasten the gold components. The upper and lower sections both have an embossed scrolling tendril design, the central band is openwork with S-spirals bordered in filigree wires. The base has a simple eight-petalled rosette design. The vessel recalls the famous golden bowl from the Celtic burial at Schwarzenbach (Germany), and the gold foil from a drinking horn found at Eigenbilsen (Belgium), both dating to the late fifth century BC. However, the present vessel has a more obvious Mediterranean influence in the tendril scroll, and the use of filigree rather than embossing alone demonstrates its more sophisticated craftsmanship. The S-scrolls of the central band can be parallel in Celtic metalwork, but the scrolling tendrils are more clearly influenced by Mediterranean forms. The use of filigree on the bowl is a strong indication of such direct influence. Perhaps the nearest technical parallel to the present object is a bronze helmet with an applied embossed and openwork gold cladding from Agris, France, of the fourth century BC. This has scrolling tendrils somewhat like the bowl and openwork bordered by beading.

The components of the vessel are of hammered gold foil. These are shaped to the form of the wooden core vessel and held in place with small, round-headed nails of gold, as can be seen in the exterior and internal photos, right. What is noteworthy is that the tendril design passes neatly over the join a sophisticated touch that must have complicated the object’s manufacture. Obviously, the wood core cannot have been in place when the tendrils were worked. The main contours of the tendril design were worked from the back of the sheet, the details added from the front. The latter consist of the sharpening up of the outline and small dotted highlights. The central openwork band is of similar thin gold foil pierced with a chisel. The S-scroll motifs are outlined in what is termed spiral beaded wire, that is a round-section wire which has been rolled under an edge at a slight angle to produce a spiralling groove along it – much like a screw-thread. Spiral beaded wire first appears in about the seventh century BC, initially in Phoenician and Etruscan goldwork, slowly disseminating to the Greek world.

The use of spiral beaded wire here is noteworthy. The ‘beaded’ borders to the openwork on the Schwarzenbach bowl, Eigenbilsen drinking horn panel and probably the Agris helmet, are embossed in the gold foil, not applied decorative wires – a common feature in Celtic gold work. Decorative wires of are very rare in Celtic goldwork. The wires were almost certainly soldered in place with what is termed diffusion soldering. A copper-based soldering substance such as copper acetate, ground malachite or a similar copper compound, was mixed with organic glue. The wires were stuck in position with this sticky soldering mixture and then the whole piece heated. The heat causes the organic glue to burn to carbon and then in the presence of the carbon the copper compound reduces to metallic copper which alloys with, and diffuses into, the surrounding gold to produce a good but unobtrusive joint. This was the usual method for joining gold jewellery components in antiquity. The main components of the bowl had to be fastened with mechanical fixings, such as the nails used here, because soldering clearly impossible once the wood core was in place.

The original wood core vessel was probably formed with a wood-turning lathe, a technique in use in Celtic Europe by the sixth century BC. When the bowl was examined, there were no traces remaining of the original wooden core vessel. However, photos were seen of the parts dismantled and the undersigned is convinced that the various components do belong together.

The composition of the gold elements are as follows:

Gold Silver Copper
Lowest section 69.06% 28.89% 1.39%
Top section 71.17% 27.30% 1.24%
Central openwork band 72.32% 24.17% 2.55%

No other elements were detected with certainty at levels resolvable by the equipment apart from a possible trace of iron which may derive from the burial environment. This overall purity and the silver to copper ratio are consistent with an ancient origin and the gold may well be native gold, not refined. The relatively high silver content means that the vessel could almost be described as being of electrum. The slightly higher copper content in the central band may be due to the soldering of the filigree which could locally enrich the copper content. No elements were detected that would cast any doubt on the authenticity of the object. As a comparison, the foil from the Eigenbilsen drinking horn was of gold with about 14% silver and 2% copper.