Crucible and furnace

Heat enough to pour

Bronze melts at roughly 950–1000 °C, depending on the tin ratio. Charcoal and a bellows can reach that temperature; a simple open hearth cannot. Everything about Bronze Age casting turns on that gap — the difference between a fire you can cook over and one you can work metal with.

The furnace, in most Bronze Age contexts, is a shallow clay-lined pit or bowl, sometimes with low walls built up around it. Structural complexity varies by region and period, but the underlying principle is constant: contain the fuel and the charge close together, reduce heat loss from the sides, and drive oxygen in from below or at the base. The clay lining matters for insulation and for chemistry — unlined soil pulls heat away too fast and introduces unwanted minerals.

Charcoal, not wood, is the fuel. Green or unseasoned wood burns too cool and too smoky. Charcoal, made in controlled reduction conditions, is almost pure carbon and burns hot and clean. Woodland management for charcoal production represents a significant and often underestimated part of the resource budget of any community smelting or casting metal, and there is good charcoal evidence at Bronze Age copper-production sites across the Alps and the British Isles, including the great mine complex at Mitterberg in the Austrian Salzburg region.

Bellows, tuyères, and the shape of air

Forced air is not optional. Without it, even good charcoal burns below the liquidus — the temperature at which the alloy flows freely. The standard Bronze Age solution is a skin or hide bellows, pumped by hand, feeding air through a clay tube. That clay tube — the tuyère — is the most archaeologically visible part of the system. Tuyères survive as heat-vitrified fragments in excavated metalworking deposits; their angles and diameters tell us how air entered the fire and at what approximate velocity. At Flag Fen itself, and at contemporary metalworking sites across temperate Europe, tuyère fragments appear in exactly the kind of contexts — alongside slag, crucible fragments and fuel ash — that mark a casting or smithing floor.

The crucible sits in or immediately above the burning charge, surrounded by fuel, not resting on a grate over it. Proximity to the fuel is everything. Most Bronze Age crucibles are small, triangular or rounded vessels of coarse tempered clay, purpose-made for metalworking. They survive as cracked, slag-encrusted sherds, often with a telltale copper or tin stain on the interior that laboratory analysis can identify precisely. Some show pour spouts or deliberate shaping at the lip; many show repeated use evidenced by successive layers of vitrified slag. A single crucible typically held no more than a few hundred grams of metal — enough for a small axe or several smaller castings.

Getting the charge to temperature is one problem; holding it long enough to pour cleanly is another. Bronze begins to oxidise at the surface once molten, forming a skin of cuprite that will ruin the casting if it folds back in. Working quickly and knowing when the melt is ready — colour, fluidity, the behaviour of the surface — are tacit, embodied skills that cannot be read off any artefact. Experimental work by archaeometallurgists, including the replicated Bronze Age casting experiments documented by groups working in the tradition of pioneering researchers such as Peter Northover, makes clear that the margin between a good pour and a failed one is narrow. Temperature, timing, and the control of oxidation all matter within ranges a skilled caster learns to read but cannot easily explain.

Plate iii

Flux — usually charcoal dust or a small addition of another mineral — could be used to protect the melt surface and help impurities float clear, though the evidence for specific Bronze Age flux practice is inferential rather than direct. What is direct is the evidence of failure: miscast ingots, incomplete fills, and the cold shuts (surface seams formed when two fronts of cooling metal meet) visible on poorly executed finished objects.

Green or unseasoned wood burns too cool and too smoky.

The relationship between the alloy proportions in the final object and what the caster put in the crucible is never simple. Tin is volatile at high temperatures and some is always lost to the atmosphere; lead, sometimes added in small amounts, lowers the melting point and improves flow. The crucible was not just a vessel — it was a reactive environment, and the caster who controlled it was managing chemistry without knowing that word.