Silver and Silver Working

This is the next in our series of posts on metal-working written by Dr. Kevin Leahy, PAS National Finds Adviser. The articles were first published in The Searcher magazine and are reproduced here with kind permission of Harry Bain, editor for The Searcher.

The Mildenhall Treasure is one of the high points of British archaeology. Apart from the artistry and the classical decoration, it is worth reflecting on the skill of the 4th century metalworkers who raised the repousse from flat sheets of silver. Copyright: The Trustees of the British Museum.

Of the 1.4 million objects recorded by the PAS to date, almost 180,000 (around 12%) are made from silver, representing a respectably high percentage. However, 78% of these are silver coins, either as single finds or as part of coin hoards which, while greatly adding to the total, are dealt with as groups to keep things in proportion.

Along with gold, silver is one of the “noble” metals; the other noble metals such as platinum were not known in antiquity and don’t concern us. These metals are known as noble because, unlike base metals, they don’t corrode – although silver does have its problems. I once excavated an Anglo-Saxon silver toilet set which, on first appearance was silver but, as we watched, it turned violet, then purple, then black. This was a photo-chemical reaction as the silver chloride covering the tweezers reacted with sunlight. Most disappointing. Silver is a lovely metal; bright, lustrous and highly workable, it is both easily cast and wrought giving craftsmen a superb medium on which to apply their skills.

Finds from around the Aegean Sea show that silver was being used as early as the 4th millennium BC but we have no evidence for its use in Britain before the later Iron Age, around 100BC, when we then see a substantial use of silver coins (there are over 17,000 Iron Age coins on the PAS database).

The Romans loved silver and some of their work, both in coins and tableware, has never been surpassed. Claims (perhaps exaggerated) of its mineral wealth may have played a part in the Roman’s decision to conquer Britain.

1. Only a few of the thousands of Roman brooches recorded by the PAS are silver. This gilded example (LVPL-80B2F2) is of “Polden Hill” type and is identical in form to many of the copper alloy brooches we see. 2. A Roman dress pine from Leicestershire in the form of a woman’s hand holding a scent bottle, AD50-200 (NARC-2D8651). 3. There has been some debate over the dating of this cheerful silver-gilt bird from Lincolnshire, but an Early Medieval date seems most likely, AD700-800 (NLM-78C696). 4. A Middle Anglo-Saxon  silver pin with the typical gilding below the head, from Lincolnshire and dated AD700-900 (LIN-4C7679). 5. A Middle Anglo-Saxon silver-gilt pin head from Lincolnshire, AD750-800. Up to the 8th century gilding was commonly applied to silver objects. This pin head formed part of a set of linked pins (NLM-028751). Copyright: Portable Antiquities Scheme, License: CC-BY.

Most silver is found in combination with lead. Lead is easily extracted from its ore but separating the silver from the lead was laborious and time-consuming.  It was done by a process known as cupellation where the lead was placed in a shallow hearth at a temperature of around 1000-1100ºC and a blast of air played over it. This oxidised the lead converting it to litharge (lead oxide) which was skimmed off or absorbed by the bone ash that lined the hearth. The silver, unaffected by the air blast, remained untouched as a pellet in the middle of the hearth. 

This sounds simple but imagine treating tons of lead in this way. The litharge was then re-smelted to turn it back to lead, which was a by-product. Over-production meant that the market was saturated with more lead than anyone needed.

The trouble here is the silver content of most British ores is disappointingly low and it’s not really worth the trouble extracting it. The Derbyshire ores contained around 2 to 6 ounces per ton which stands in stark contrast to the great silver mines at Laurion, which paid for the ships that defended Athens, and contained about 600 ounces of silver per ton.

The Romans seem to have limited themsleves to ores containing more than 20 ounces of silver to the ton of lead. Despite claims made on Roman lead ingots that they are “EX ARG” or “EX ARGENT” (from the silver mines), analysis of ingots showed that they still contain a low level of silver – no attempt had been made to extract the silver as it was simply not worth the trouble. But why are British ingots marked “EX ARG BRIT” when argentium (silver) wasn’t there? Was something dodgy going on? It all seems rather odd.

6.  A Middle Anglo-Saxon strap end decorated with Trewhiddle style inlaid with niello, AD800-900 from Staffordshire (WMID-4462B7). 7. Silver ingots like this are difficult to date but they are found in Viking hoards. This one was found in Essex and is dated to AD800-1000 (BH-7253E7). 8. This brooch was found in Norfolk as part of a hoard of 23 coins, four brooches and two strap ends. This fragment of a disc brooch shows the use of niello inlay (NMS-972E58). 9. Heavy stamp marks like these are a feature of many pieces of Viking silver (ESS-9CB5B8). 10. This silver ingot fragment from Lincolnshire shows the single chisel cut often seen on Viking ingots (NLM-1B15F9). Copyright: Portable Antiquities Scheme, License: CC-BY.

Pure silver melts at 962ºC but was rarely used as such: adding other metals to silver to form an alloy reduces its melting point. Sterling silver, the most commonly used silver alloy, contains 7.5% copper and melts at 893ºC. Not only is the melting point reduced but Sterling silver is harder and stronger than the pure metal. This lowering of the melting point is also the principle behind solder where an alloy of two parts silver to one part copper will have a much lower melting point than pure silver making it possible to join pieces together without risk of melting them. The surfaces to be joined were cleaned carefully and the solder applied along with borax which acted like a flux preventing the surfaces from oxidising. When heated, the solder flowed into the joint making a strong and almost invisible bond. Soft solder, an alloy of lead and tin has a melting point of around 300ºC and was much easier to use, but after burial objects can fall apart.

The watlington Hoard found in Oxfordshire contained 186 coins, 7 items of jewellery and 15 silver ingots (SUR-4A4231). Copyright: The Trustees of the British Museum.

Silver  is easily cast. The method used to produce many small objects was the “cire purdue” (lost wax) process, a method so clever and ingenious that it still invites wonder. The first step was to make a model of the required object using beeswax, a material that is soft and easily worked. Once this was done the necessary channels for the metal to flow into the mould were added, also made of wax. The whole thing was then covered in clay. To ensure that the full detail was captured this would be done by repeatedly dipping the wax model into a clay slurry to build up the mould thickness. When this was dry, it was turned upside down and heated in a kiln. This fired the clay and melted the wax which ran out to leave a hollow mould which was filled with molten silver. Intricate objects with fine detail could thus be easily made.

Silver is also highly malleable and the ease with which it could be shaped by hammering led to the production of the plate and table vessels that adorned Roman and Medieval aristocratic tables. Breath-taking hoards of Roman silverware have been found such as the Mildenhall TreasureSilver vessels can be made by “sinking” in which a disc of silver is hammered into a hollow cut into a block of wood. Soft and malleable, silver can be cold worked but, as work progresses it becomes increasingly work-hardened so it is necessary to “anneal” it – heating it up to soften it and prevent it from cracking. This may need to be done repeatedly. After “sinking” the silver vessel was placed over a round-ended iron stake from planishing – using concentric light taps from a smoothly polished hammer to further shape the object.

11. A silver brooch or badge made from a penny of Edward the Confessor, found in Hampshire. These brooches are an odd phenomenon as only one face was gilded (usually reverse) and fixings look too weak for use as a brooch (WILT-C94353). 12. A Medieval silver seal matrix bearing the figure of St. Peter and the inscription + SIGILLVM PETRI DE BARASTRE. An American professor contacted us to identify the man as Peter de Barastre, a canon of Wherwell Abbey (HAMP-4E0317). 13. A silver gilt Medieval finger ring found on the Isle of Wight (IOW-9F11B3). 14. A gilded Medieval brooch found in Worcesterhire, AD1270-1350 (WMID-EF7056). 15. A Medieval silver-gilt finger ring bearing the inscription “AVE MARIA” – a popular inscription as the Virgin Mary was greatly venerated (NMGW-115DBC). Copyright: Portable Antiquities Scheme, License: CC-BY.

Repoussé decoration was much used to decorate silver plate. This involves the design being raised by hammering. One can imagine the difficulties of hammering out fine decoration on a thin sheet silver plate. The answer was ingenious: the top of the surface to be raised was covered in a thick layer of pitch and, working from beneath, the design was hammered down into the resilient pitch. When complete the pitch could be sharpened up by “chasing” using fine, flat-ended tools to hammer down from the upper surface.

Silver could be further embellished by gilding or by the application of niello or enamel. Gilding was carried out by dissolving gold into mercury to form an amalgam which was painted onto the surface to be treated. The object was then heated so that the mercury boiled off, leaving a thin, fine layer of gold on the treated surface. Niello was a black silver sulphide which was applied to the surface of silver to give a bold contrast.

16. A small silver-gilt Post-medieval dress hook found in Buckinghamshire (BUC-C09A34). 17. Silver gilt dress hook decorated with filigree wire and granulation both fused onto the surface of the silver. Once rare, metal-detecting has shown that these fittings must have been relatively common during the Tudor period (IOW-6CD1A5). 18. At one time, pins like this one were considered to be Viking but they are now known to be Post-medieval, AD1500-1600 (KENT-CE8D6B). 19. Post-medieval pin found in Surrey (SUR-C52692). 20. A silver spur, probably for attaching to a fighting cock. Not a nice object but still interesting (SWYOR-B06054). Copyright: Portable Antiquities Scheme, License: CC-BY.

While gold is the premier metal, I often think that we see a higher level of craftsmanship on silver objects where, using a workable material and not having the WOW factor of gold, the artisan had to try that bit harder.