Never Reclaimed: Why Was a Hoard of Copper Crosses Buried Beneath a Tree in Malawi?

It's 1967, and you're a farmer in Chombe, Malawi, happily digging at the base of a tree when your shovel hits something hard - a stone, perhaps? But then you look closer; it's shaped like an X and made of metal. As you dig, 7 other copper Xs follow; you keep them for a few years before showing them to Paul Cole-King, an archaeologist surveying the area.

These are copper ingots, and hundreds have been found in Zambia and Zimbabwe, but your little hoard is different: it's lighter, rougher, and something just isn't quite right. But what?

Now, new research has reanalysed these ingots and found where they came from and how they got there, and the answer places Malawi firmly in a vast and sophisticated Iron Age commercial world. One where copper travelled over 1000km along the Kafue River, east along the Zambezi and finally overland into Malawi. It's not random wandering but an engineered supply chain taking the longer but far easier route, one that would make any modern businessman proud. And it's a route remembered in Chewa oral traditions, which tell of the Maravi state founders as migrants from the same copper-rich area as the ingots.

But someone, somewhere along the way, made another decision: stop, remelt, and multiply the copper into thinner, lighter versions. Since none of the ingots has mixed sources, they likely remelted existing ingots one at a time, though they may have remelted scraps into new ingots. Yet the mystery lingers: who recast these ingots, what did their journey look like, and why were they buried beneath a tree at Chombe, never to be reclaimed?

People of Copper

There are many people and regions connected to copper but for this study we only need to worry about three of them. The first is the Copperbelt on the Zambia-Democratic Republic of the Congo border. Even today it is home to the world's largest copper and cobalt reserves; even the uranium for the Manhattan Project came from there (that top-secret US project that created the world's first nuclear weapons). Colonial accounts from ~1900 describe giant mining trenches up to 1.5 km long and 45 m deep on the Copperbelt, and 200 m-long pits at the Kipushi ore body near Lubumbashi, DRC. Organised states such as the Luba and Lunda lived alongside smaller communities there. According to lead author Jay Stephens,

"Copper seemed to be the major connector for these communities, as we have now identified copper from the Copperbelt across the broader region of southern Africa, including at Ingombe Ilede and in Malawi."

Which brings us to the second copper-linked community, the Maravi. Maravi was a precolonial state on the southern shore of Lake Malawi famed for its ivory and ironwork, linked to trade networks from the Copperbelt to the Indian Ocean.

"One interesting aspect of the Maravi is that their oral histories link the founding populations of the state to a migrating community from the Copperbelt."

The final region is Ingombe Ilede; no massive state but a burial site on the Zambia-Zimbabwe border where the dead were buried with anvils, hammers, tongs and X-shaped copper ingots, perhaps the graves of the traders who once carried copper across the land.

Crucially, Malawi has no copper deposits, and the few deposits in Mozambique were little used. So exactly how far did this copper travel?

Copper, Isotopes, and Sports Cars?

Eight ingots were recovered in 1967; in 1970, the Malawi Department of Antiquities acquired them. Since then, three have gone missing. So Stephens and colleagues made do with the 5 survivors. It is a sobering reminder of how fragile heritage is: these ingots survived centuries underground, yet three went missing within fifty years of being found. How much heritage have we lost simply by finding it? But if we never find it, it may just as easily remain forgotten.

The design hints at a 14th- to 18th-century age, but more precise dates are not possible, and the radiocarbon date from the 70s is questionable. However, it too hints at a similar period.

Looking closely revealed they were sloppily made, like seriously. Decorative markings were blurred, and their rough surfaces were pockmarked by craters from steam bubbles bursting through the molten metal.

How did this happen, and are ingots elsewhere equally sloppy? No. The hoard under the tree is the odd one out.

Someone seemingly pressed a normal-sized ingot into damp sand to make a mould, hence the bubbles, with two centre marks even improvised by hand before casting.

"Rather than using clay or soapstone molds, which impart crisp details to cast metals, we identify that ingots in the [Chombe hoard] were cast in sand molds," Stephens explained. "While this is faster to produce... it cannot translate details to a cast object as well and can impart imperfections if used while still damp. These issues are all visible in this cache of ingots, and suggests that someone less experienced was quickly trying to reproduce them."

Where did the metal come from? That's where isotopes come in; you've heard isotopes work like fingerprints, but how? Stephens uses an interesting analogy. Isotopes are versions of an element with slightly different masses, like a bus and a sports car: both are vehicles, but differ in weight.

A mass spectrometer exploits those masses like a racecar track: lighter isotopes hug the inside of the curve and zoom past. The heavier ones swing wide, letting the machine measure how many heavy and light isotopes a sample holds. Copper sourcing uses lead isotopes, and lead isotope mixes depend on how long ago their ore deposit formed. The DRC's Copperbelt ores developed at different times than, say, those in the Magondi Belt in Zimbabwe and thus have different signatures.

So, based on our isotope race cars and buses, where did the five ingots come from? All five point to the Kipushi ore deposit, with copper so pure (low in iron and nickel) that it could be smelted without producing slag (waste left over from refining and melting ores).

Although three are missing, analyses in the 1970s suggest at least two may have come from the Copperbelt, making the hoard a mix of copper.

Burying, Forgetting and Reanalysing a Copper Hoard

The analysis revealed the Chombe hoard was Kipushi copper, roughly 800km away as the crow flies. But people don't fly, and a straight line means crossing rugged terrain, mountains, valleys, and rivers; a tough journey.

So how did they get there? Two possibilities.

One, they truly travelled overland, possibly on the same routes as the Bisa traders of ~1760 who carried ivory, slaves, and copper from the Copperbelt through Chewa territory to southern Lake Malawi, with Chombe sitting neatly on this path.

The other was a longer but far easier journey of roughly 1200 km, mostly by canoe, paddling down to Cahora Bassa.

Seems a weird choice? Not really; we ship things via ocean networks all the time, even though flying is faster. Why? It's cheaper, carries more and heavier goods, and avoids dangerous overland routes.

So perhaps these ingots once sat in a canoe as traders merrily paddled down the Kafue, turned east along the Zambezi, simply row, row, rowing their canoe, hey, that reminds me of a song.

Row, row, row your canoe, gently down the stream, merrily, merrily, merrily, merrily, trading copper is the dream…

Okay, not quite the lyrics, but fitting anyway. At some point, possibly near the now-drowned cataracts at Cahora Bassa, they would have come ashore for the final 220km to Chombe.

Remember that burial ground where people were buried with ingots, anvils, hammers, and tongs? This second route passes right by it, so perhaps the hoard marks a stop on that ancient trade corridor linking the Copperbelt to Maravi, Chombe, and beyond.

At some point, someone, perhaps a trader, perhaps an opportunist, remelted the ingots into thinner versions and buried them. Why they never came back is anyone's guess, and your imagination could run wild.

Maybe they simply forgot where they buried their treasures, like a squirrel forgets last year's nuts; perhaps they died first, or someone caught wind of their remelting and punished them, or for a happier ending, they came into wealth and did not care for 8 measly ingots. The possibilities are endless, and sadly all unprovable.

So the farmer's find finally has answers: the ingots were lighter and rougher than any of their kind, copies of the originals pressed into moist sand and recast. They came from hundreds of kilometres west, from the mines of Kipushi, likely carried by canoe along rivers before being lugged overland for the final stretch. This was no random path, but a well-thought-out supply route tackling the challenges of overland travel. These people were solving the same problems every trader solves today: how do you move something heavy and valuable over difficult terrain at lower cost and risk?

Even the recasting was, in a way, a business decision: more units from the same raw material. " [Today]We call it value addition and market penetration," said Smeda Matundu, chief curator of Malawi's Chichiri Museum.

The find lends weight to what Chewa oral traditions have said all along: the founders of the Maravi state came from that copper-rich land, where identity, trade and political power were intertwined. Old tales and folklore often hold a grain of truth; sometimes it just takes science a bit longer to catch up and figure out which grain it was. Even the most (seemingly) humble artefacts can have big stories to tell.

Why do you think the ingots were never reclaimed, and was the remelting a smart business decision or something else entirely and do you think old tales and oral histories deserve more weight in how we reconstruct the past?

Let me know what you think, and if you'd like to support this blog and the person writing it, you can chip in once, monthly, or yearly. Only if you'd like to.

Reference

Stephens, J., Renson, V., Killick, D., Kaliba, P., & Matundu, S. (2026). Long distance trade of copper ingots from the eastern Copperbelt (Democratic Republic of Congo) to central Malawi. Journal of Archaeological Science: Reports, 73, 105845. https://doi.org/10.1016/j.jasrep.2026.105845