By Emilia Wellesley · Published May 1, 2024 · Updated September 12, 2026
Last reviewed: September 12, 2026.
What Is the Baghdad Battery?
The Baghdad Battery is a clay jar, roughly thirteen centimeters tall, holding a copper cylinder and an iron rod, recovered near Baghdad in the late 1930s and dated by most scholars to the Parthian or early Sasanian period (roughly 250 BCE to 650 CE). Wilhelm König, who studied it at the Iraq Museum in Baghdad, asked in 1938 whether it once produced electric current. Later scholars argued it held written blessings or incantations. The artifact still resists a single answer.
A few objects in archaeology refuse to settle into the category their finders chose for them. The jar from Khujut Rabu is one of them. It is a small, ordinary-looking pot with extraordinary insides: a sheet of copper rolled into a tube, an iron rod corroded almost to a stub, and traces of a residue read as either bitumen seal or chemical electrolyte. For nearly a century, archaeologists, museum curators, electrical engineers, and television crews have argued about whether the assemblage represents the world’s oldest galvanic cell or simply a clever piece of religious storage. The dispute is genuinely unresolved, and the reasons it remains unresolved are themselves a useful lesson in how ancient material is read.
This guide walks through the discovery, the competing reconstructions, the experiments that tried to settle the question, and the wider cultural argument the artifact has come to anchor. The aim is to hold both the König hypothesis and the scroll-container reading in view without flattening either, situating the jar within the broader landscape of historical and archaeological mysteries.
Discovery at Khujut Rabu
The jar was uncovered at the village of Khujut Rabu, just outside Baghdad, in 1936. No two accounts of the find agree: some describe an excavation, others say König came across the objects in the museum’s collections. They ended up in the National Museum of Iraq, where the Austrian-born painter and archaeologist Wilhelm König studied them. König noticed that each jar contained a small assemblage of metals that looked, to his eye, less like grave goods than like apparatus.
The Object Itself
The principal artifact is a buff-yellow ceramic jar approximately 13 to 14 centimeters in height. Inside sits a copper cylinder, formed from a rolled sheet roughly 9 centimeters tall, soldered along its seam with a lead-tin alloy. A heavily corroded iron rod, suspended down the cylinder’s axis, projects slightly from the top. A plug of asphalt or bitumen seals the rod and the cylinder’s mouth, isolating both metals from the surrounding clay. The interior contains traces of a corrosive residue. König read this assembly as electrochemistry. Later analysts read it as preservation.
Dating Disputes
König placed the jars in the Parthian era, roughly the late centuries BCE through the early third century CE [1]. St John Simpson of the British Museum has since questioned that date: “Although this collection of objects is usually dated as Parthian, the grounds for this are unclear,” he told the BBC in 2003, adding that “the pot itself is Sassanian,” which would place it after 224 CE. The jars have never been radiocarbon-dated; the dating remains stylistic and stratigraphic.
The König Hypothesis: Galvanic Cell
In 1938, Wilhelm König published a two-page note in the Berlin journal Forschungen und Fortschritte asking whether the Khujut Rabu jar might have functioned as a primitive galvanic cell [2]. He drew the analogy to the Voltaic pile that Alessandro Volta announced in 1800: two dissimilar metals separated by an electrolyte, generating a small but measurable voltage. The copper cylinder and iron rod fit the geometry, the bitumen plug fit the role of an insulator, and a sour electrolyte such as wine, vinegar, or grape juice would, in principle, complete the circuit. König speculated that Parthian craftsmen used the cells to electroplate gold onto silver, anticipating modern electrochemistry by nearly two millennia.
Why the Idea Took Hold
The hypothesis arrived in print on the eve of the Second World War, but the argument travelled. Willard F.M. Gray, an engineer at the General Electric High Voltage Laboratory, built a replica using copper sulphate that, in Willy Ley’s 1954 phrase, “worked quite well for a short time”; no voltage from his tests survives in the primary record. In 1978 Arne Eggebrecht, then director of the Roemer-Pelizaeus Museum in Hildesheim, connected many replicas filled with grape juice and claimed a thin silver deposit, though the museum later told the BBC that no written record or photographs of the test exist.
What the Hypothesis Requires
For the galvanic reading to hold, the original jar must have contained an electrolyte at burial or immediately before, and the copper-iron geometry must have been the operator’s deliberate aim rather than a side-effect of another use. Both points are contested. No surviving residue has been chemically identified as wine or vinegar; the corrosion patina is consistent with a range of mildly acidic environments, including soil leachate during burial. The geometry, as the British Museum would later argue, also matches a quite different purpose.
The Scroll-Container Reading: Blessings and Incantations
The most developed counter-argument came from Emmerich Paszthory, who argued in 1989 that such jars were containers for blessings or incantations written on organic material, a reading the conservation chemist Gerhard Eggert judged in 1996 to be “much more probable” than the power-source claim [3]. On that account the metal parts held or protected a rolled text and the bitumen plug sealed it. Paul Craddock, a metallurgist at the British Museum, took a third line: he told the BBC in 2003 that gilding from the region and era is “conventional gild plating and mercury gilding,” yet he still thought the objects were batteries, perhaps hidden in temple statues to give worshippers a mild shock.
Comparable Finds
König himself had recorded the comparable finds: bronze cylinders from Seleucia-on-the-Tigris that held rolled papyrus, and rolled bronze sheets from Ctesiphon. Those parallels, rather than any galvanic design, are what the scroll-container reading rests on, and they suggest the Khujut Rabu cluster may be a small consecrated deposit rather than a workshop cache.
Why Galvanic Use Is Hard to Square
If the jars were batteries, one would expect connection points: external terminals, conductive wires, a circuit. None have been recovered. One would also expect them to occur at workshop sites near goldsmiths, not in apparent grave deposits. And the operator would need a body of related practice — texts, training, a guild memory — and no Parthian or Sasanian source describes anything resembling controlled electrochemistry. The galvanic reading must postulate a knowledge tradition that left no other trace.
What the Replicas Showed
A century of replicas has produced a stable result: a single Khujut Rabu-style jar, charged with a mildly acidic electrolyte, generates between roughly 0.5 and 2 volts. That is enough current to gild a small surface, to deliver a faint tingle to a moist hand, or, in series, to power a low-current LED. Whether the ancient operator used any of these effects is a separate question.
MythBusters and the Television Test
In 2005, the Discovery Channel program MythBusters built ten replica jars, charged them with lemon juice, and wired them in series. The set measured roughly four volts, enough to plate a small copper token with zinc overnight; five jars could push a painful current through acupuncture-type needles inserted in the skin, though all ten could not shock dry skin [4]. The experiment confirmed feasibility but left intent untouched. Engineers can build a battery from the surviving form. The question is whether ancient craftsmen ever did.
The Eggebrecht Gilding Test
Eggebrecht’s 1978 demonstration, staged for journalists during an exhibition, is said to have left a thin silver deposit using a series of replica jars and a grape-juice electrolyte; later retellings turned it into gold, and the museum has no notes or photographs of it. Modern conservators, however, have argued that the gilding observed on Parthian silver is almost universally consistent with mercury fire-gilding, the standard hot-amalgam technique attested in surviving recipes and ethnographic sources. No Parthian or Sasanian object has been independently identified as electrogilt.
Every Replication of the Baghdad Battery Since 1938
Six attempts to rebuild the Khujut Rabu jar are on record between 1938 and 2026, and they agree on one point: copper, iron, and a sour or caustic liquid will produce a voltage, with the figure depending entirely on what the rebuilder poured in. On the documentary record: the first claim was not a measurement at all. Wilhelm König’s two-page note in Forschungen und Fortschritte (vol. 14, no. 1, 1938, pp. 8-9) carried a question mark in its title, “Ein galvanisches Element aus der Partherzeit?”, and ventured only that “from its parts and their arrangement one might think that it must be a kind of galvanic element or battery.” In the same note König recorded the papyrus-bearing bronze cylinders from Seleucia and the rolled bronze sheets from Ctesiphon, so the scroll-container comparison (see The Scroll-Container Reading) was on the table in 1938, placed there by the man who proposed the battery.
Gray, Eggebrecht, and the Television Jars
Only the documentation of the three replicas described above (Why the Idea Took Hold, The Eggebrecht Gilding Test, MythBusters and the Television Test) is added here. Willard Gray’s General Electric model used copper sulphate and, in Willy Ley’s 1954 phrase, “worked quite well for a short time”; the chemist Gerhard Eggert, writing in Skeptical Inquirer (May/June 1996), locates Gray’s surviving models at the Berkshire Museum in Pittsfield, Massachusetts, but gives no voltage, and I could not trace the half-volt figure of later summaries to a primary report. Arne Eggebrecht’s 1978 gilding demonstration was staged for journalists during the Hildesheim exhibition Sumer Assur Babylon. Asked by the BBC’s Arran Frood in 2003, the museum researcher Bettina Schmitz replied that “there does not exist any written documentation of the experiments which took place here in 1978,” not even photographs (BBC News, 27 February 2003); the same report describes the claimed deposit as silver, one ten-thousandth of a millimetre thick, so the gold-plating version belongs to later retellings. The MythBusters build of 23 March 2005 remains the only well-documented series test, at roughly four to four and a half volts from ten lemon-juice jars, depending on the account.
The Two-Cell Reconstruction of 2026
The newest replica is the work of Alexander Bazes, an independent researcher in Westchester, New York, published as Sino-Platonic Papers no. 377 (January 2026), a University of Pennsylvania series for unconventional work, and reported by Tom Metcalfe in Chemistry World on 23 January 2026. What the 2026 paper actually establishes: the copper-and-iron “inner cell” alone gave Bazes about 0.26 volts with salt water and 0.35 with lemon juice, as feeble as every earlier replica. His new step was to treat the unglazed clay jar and the tin-lead solder on the copper tube as a second cell, a tin-air battery, by filling the jar with lye (potassium hydroxide). That outer cell alone read over 1.1 volts, and the assembled device delivered over 1.4 volts, enough to raise bubbles on a copper plate in a salt bath and etch it overnight. Bazes reports no plating test, and his paper does not engage Gray, Eggebrecht, or the scroll-container reading.
The Replication Table
Every voltage below comes from a replica; no source opened for this update records a measurement taken on the original jar.
| Year | Who | Electrolyte | Volts measured | What it powered | Documented? |
|---|---|---|---|---|---|
| 1938 | Wilhelm König, Iraq Museum | “Acid or alkaline liquid” (conjectured) | None measured | Nothing; a hypothesis | Yes: 1938 paper, 1940 memoir |
| 1940s | Willard Gray, General Electric | Copper sulphate | Unstated; ~0.5 V in later summaries | Nothing recorded | Partly: Ley 1954, Schwalb 1957 |
| 1978 | Arne Eggebrecht, Hildesheim | Grape juice (BBC 2003) | Not recorded | Claimed thin silver deposit; gold in retellings | No: no notes or photographs found (2003) |
| 2005 | MythBusters, Discovery Channel | Lemon juice, ten jars in series | ~4 to 4.5 V total; ~0.4 V per jar | Zinc-plated a small copper token overnight; current felt through inserted needles | Yes: broadcast 23 March 2005 |
| 2026 | Alexander Bazes, Sino-Platonic Papers 377 | Lemon juice or salt water (inner); lye (outer) | ~0.35 V inner; >1.1 V outer; >1.4 V combined | Bubbling and etching of a copper plate | Yes: photographed readings, published paper |
What the Replications Do and Do Not Prove
A replica that lights a meter proves that the geometry can hold a cell, not that anyone in Parthian Mesopotamia built one for that reason, and every serious critic since Gerhard Eggert in 1996 has rested the case on exactly that distinction. Eggert put it in one sentence: successful experiments “can show only a supposed ancient technique to be possible, but never its application.” Two things get conflated here: the electrochemistry, which nobody disputes, and the intent, for which the evidence is what it was in 1938. No Parthian or Sasanian object has been shown by analysis to be electrogilt, no wire or terminal has appeared, and the “secret” gilding method of Baghdad silversmiths that König cited turned out, on Eggert’s tracing, to descend from John Wright’s Birmingham process of 1839, ninety-nine years old when König wrote. The caution runs both ways: C. MacKechnie Jarvis suggested in 1960, without seeing the jar, that it might be a nineteenth-century telegraph cell, the same worry that hangs over the Coso spark plug.
Where the Scroll-Storage Reading Stands in 2026
Bazes’s 1.4 volts raises the ceiling of what the jar could have done, but it has not moved the archaeologists who work on the period. William Hafford, a curator at the Penn Museum, told Chemistry World that the battery idea “is not persuasive” and read the related finds, including one clay jar holding ten nested copper vessels, as ritual deposits: a prayer or curse dropped through the neck, sealed with bitumen, and buried for the deities of the underworld. The scroll-container reading, its purpose now made explicit, has an unlikely ally. In a guest post on Victor Mair’s Language Log (8 January 2026), Bazes himself proposed that the current’s purpose “may have been solely to corrode the iron rod,” releasing the text wrapped around it; on that account a working battery and a magical container are the same object. The tally for 2026 is therefore this: the physics is settled higher than before, the intent is as open as it was, and the storage reading remains the one that requires no undocumented craft tradition, the test this site applies to the Dorchester Pot.
Where the Question Sits Now
The professional consensus among archaeologists currently leans toward the scroll-container reading, primarily because it requires no undocumented knowledge tradition and matches comparable finds from the same region. The galvanic reading retains a smaller community of defenders: Paul T. Keyser argued in the Journal of Near Eastern Studies in 1993 that the jars were genuine cells used for analgesia, and the historian of science Marjorie Senechal told the BBC in 2003 that they “may have been batteries because they do work” [5]. Among archaeologists the verdict is blunter still: Elizabeth Stone, a specialist in Iraqi archaeology, told NPR, as HISTORY reported, “I don’t know anybody who thinks it’s a real battery in the field” [6].
The Looting of 2003
The original jars were housed in the National Museum of Iraq in Baghdad. During the looting that followed the 2003 invasion, some 15,000 objects were taken from the museum’s galleries and storerooms, and by 2018 only about 7,000 had been recovered. Many of the original Baghdad Battery artifacts disappeared in the looting, and the whereabouts of most remain unknown, so what survives for study is photographs, published descriptions, and replicas [7].
Why It Still Matters
The Baghdad Battery has become a touchstone in arguments about so-called out-of-place artifacts and ancient advanced technologies. Some popular writers treat it as proof of suppressed pre-modern science. The careful position is humbler. The jar is one piece of late antique apparatus whose function is genuinely uncertain, whose context was poorly recorded by mid-twentieth-century standards, and whose primary form does not, on its own, decide between two coherent reconstructions. Both possibilities deserve to be held openly until new evidence arrives.
How to Think About the Jar
A useful rule of thumb in ancient material studies is to ask what the artifact requires from its surrounding world. A galvanic cell requires a knowledge tradition, a workshop context, and a downstream use. A scroll container requires a ritual context, a related class of finds, and a textual culture that produces things worth sealing in metal. The second list is documented. The first list is not. That asymmetry is why most working archaeologists treat the storage reading as the safer default while leaving room for the galvanic possibility to be revived if a workshop find or a chemical residue forces the question open again.
What Would Settle the Question
Three discoveries would tighten the argument materially. A Parthian or Sasanian text describing controlled electrochemical practice would tilt the case toward König. A residue analysis showing organic decomposition products inside an unaltered jar would tilt it toward the scroll-container reading. A workshop cache pairing the jars with conductive wires, terminals, or partly gilded objects would settle it almost on its own. None of these has yet appeared. The patient discipline is to wait and to keep both readings legible in the meantime.
Reading the Object Carefully
The Baghdad Battery is sometimes pressed into service as a hero of suppressed history and sometimes dismissed as a museum-shop curiosity. Neither role does the artifact justice. It is, more modestly, a small assembly of clay, copper, iron, and bitumen whose original purpose its makers did not record and whose excavators recorded only partly. What survives is the jar itself, a handful of analogous finds, and a careful argument about how to read late antique apparatus when the surrounding documentation is thin. The honest position is to name the two reconstructions, to describe what each requires from the rest of the world, and to wait for the evidence that would prefer one over the other. The past has been waiting longer than this. It can wait a little more.
Frequently Asked Questions
The questions readers most often bring to this page concern the artifact’s alternative name and the voltage it could actually produce, and the two short answers below draw only on the sources already cited in the sections above.
Why is it also called the Parthian battery?
König assigned the jar to the Parthian period from the layers at Khujut Rabu, so his 1938 title spoke of a galvanic element “from the Parthian age,” and Paul Keyser’s 1993 paper in the Journal of Near Eastern Studies (vol. 52, no. 2) kept the phrase “Parthian galvanic cells.” The name survives even though St John Simpson of the British Museum observed in 2003 that the pot itself looks Sasanian; see Dating Disputes above.
How many volts does the Baghdad Battery produce?
A replica of the copper-and-iron cell with fruit juice or salt water gives roughly 0.3 to 0.5 volts; ten such jars in series gave MythBusters about four volts in 2005; and the 2026 two-cell reconstruction, adding lye in the outer jar, gave just over 1.4 volts from a single device. The original artifact has no recorded measurement.
Sources
- Gerhard Eggert, “The Enigmatic ‘Battery of Baghdad’,” Skeptical Inquirer, May/June 1996, pp. 31–34 (PDF); Arran Frood, “Riddle of ‘Baghdad’s batteries’,” BBC News, 27 February 2003.
- Wilhelm König, “Ein galvanisches Element aus der Partherzeit?,” Forschungen und Fortschritte 14, no. 1 (Berlin, 1938): 8–9, as cited in Alexander Bazes, “The Baghdad Battery,” Sino-Platonic Papers 377 (January 2026), and in Eggert (1996), above.
- Gerhard Eggert, “The Enigmatic ‘Battery of Baghdad’,” Skeptical Inquirer, May/June 1996, pp. 31–34, discussing E. Paszthory, “Electricity Generation or Magic?,” MASCA Research Papers in Science and Archaeology 6 (1989): 31–38.
- “Myth: Did Ancient People Use Batteries?,” MythBusters Database, Discovery Channel (archived 2013); MythBusters, “Cooling a Six-Pack,” Discovery Channel, first aired 23 March 2005.
- Paul T. Keyser, “The Purpose of the Parthian Galvanic Cells: A First-Century A.D. Electric Battery Used for Analgesia,” Journal of Near Eastern Studies 52, no. 2 (1993): 81–98; Arran Frood, “Riddle of ‘Baghdad’s batteries’,” BBC News, 27 February 2003.
- María de los Ángeles Orfila, “Could Ancient Mesopotamians Have Built a Battery?,” HISTORY, 14 July 2026.
- María de los Ángeles Orfila, “Could Ancient Mesopotamians Have Built a Battery?,” HISTORY, 14 July 2026; Craig Barker, “Fifteen years after looting, thousands of artefacts are still missing from Iraq’s national museum,” The Conversation, 9 April 2018.
