The Baghdad battery: an ancient power cell?

There was no night for this one. There was a shelf. The file begins in the storerooms of the Iraq Museum in Baghdad, where a small clay jar sat among the finds. According to König's later account[konig-1940], it had come out of the ground in 1936 at Khujut Rabu, a site near Baghdad. I have not read that memoir in the original; I know it through the review that cites it[eggert-1996], and I would rather tell you that now than later.

Scene · reconstructed from the paperworkPicture it. A pot small enough to hold in one hand. Inside it, a rolled cylinder of copper. Inside the copper, an iron rod. The diagram in this file[commons-partherjar] shows the arrangement: jar, copper sleeve, iron rod, one inside the other. The podcast account[skeptoid-640] gives the parts and their dimensions. The file doesn't say who first lifted it out of the soil, or what lay beside it. That silence matters, and we will come back to it.

What the papers do say is when it acquired a second life. In 1938 Wilhelm König, who worked at the Iraq Museum, published a short paper[konig-1938] in a German journal asking whether the object could be a galvanic element. A battery. The title ends with a question mark, and I think that is worth holding on to. It was framed as a question, not an announcement.

Even the age of the thing is not settled. König called it Parthian. The British Museum curator St John Simpson, quoted by the BBC[bbc-simpson-2003], considers the pot Sasanian instead. Both are periods long before Islam; neither is pinned down by an excavation report in this file.

So the scene, honestly told, is thin:

• a jar of fired clay, small enough for one hand (the record[skeptoid-640]);
• copper outside, iron inside (the diagram[commons-partherjar]);
• a find-spot named only in a memoir written years later (König 1940[konig-1940]);
• a museum worker with a modern eye and a question (König 1938[konig-1938]).

I keep coming back to how ordinary that list is. Nothing in it glows. Nothing in it is wired to anything. It is a container with metal inside, and a man who looked at copper and iron together and thought of the laboratory rather than the pantry. That is not foolish. Copper and iron in a vessel that can hold liquid is, on paper, the recipe for a cell. The question is whether the paper resembles the past.

The jar would spend the next decades turning from an object into an idea, and ideas leave a different kind of trail.

The documents in this file are few, and I will take them in order.

1938. König's paper[konig-1938] in Forschungen und Fortschritte proposed that the jar could be a galvanic element from the Parthian period. The original is not online; our link points to the review that cites it, and the file marks the page range as one to verify. According to the later summary of his argument[skeptoid-640], König went further than the bare suggestion and wondered whether such cells could have been used to plate small objects with metal.

1940. In his memoir of nine years in Iraq[konig-1940], König described the find at Khujut Rabu and repeated the battery idea. Again, we know it only as it is quoted, by page, in Eggert's review[eggert-1996].

1985. The first serious counter-reading in this file comes from Emmerich Paszthory in Antike Welt[paszthory-1985]. His title asks the question directly: electricity or magic? He argued that the jars served a non-electrical purpose, holding sacred scrolls or serving a ritual or storage function. We have the abstract, not the full article, and the page numbers are still to be checked.

1996. The conservation scientist Gerhard Eggert published a review in Skeptical Inquirer[eggert-1996] that went through the claim point by point and concluded the jars are not batteries. It is the backbone of this file: it carries the find date, the citations to König, and the case against.

2003. The BBC report quoting St John Simpson[bbc-simpson-2003] added a curator's voice from the British Museum: likely Sasanian, not Parthian, and doubtful as a battery. The article's exact date is still on our list to verify.

2018. The Skeptoid episode[skeptoid-640] gathered it all for a general audience: the parts, the dimensions, the 1938 paper, the reasons it would make a poor battery, and the fact that the jar was lost in 2003.

Running alongside these papers is the experiment everyone mentions. Fill a replica with an acid and it produces a small current. The dossier's own summary[eggert-1996] puts it at about a volt. That much is documented, and it is fair to say it plainly: the materials can form a cell.

There is also a visual trail. No free museum photograph of the original survives for us to use. What this file shows instead is a diagram drawn by a Commons contributor[commons-partherjar], and a drawing based on photos of the lost object. They show the arrangement. They are not the object.

And there is the modern retelling: an archaeology communicator's video[miniminuteman-yt] walking through why the battery reading fails.

Notice what is missing from the list. No excavation report. No find context. No second document from 1936. The paper trail starts with an idea about the object, not with the dig that produced it. That gap is where the turn begins.

The turn here is not a revelation. It is a slow accumulation of things that are not there.

Start with the strongest point in the battery's favour. Replicas work. Fill one with acid and you get roughly a volt (the record[eggert-1996]). But the dossier's analysis[skeptoid-640] puts the problem in one line I find hard to improve on: replica voltage shows only that the materials can form a cell, not that anyone used them as one.

Then the absences, as Eggert[eggert-1996] and the Skeptoid review[skeptoid-640] set them out:

• No wires. Nothing was found with the jar that could connect it to anything else. A cell that cannot be connected is a container.
• No terminals or conductors. The same silence, from another angle.
• No electroplated objects from the period. Gilding at the time used mercury, a fire-gilding technique that needs no current at all (the record[eggert-1996]).
• A poor design for a battery. The podcast's walk-through[skeptoid-640] explains why the arrangement would make a weak and awkward one.

Then the alternative. Paszthory[paszthory-1985] read the jars as vessels for sacred scrolls, or for a magical or storage use. That is the reading most archaeologists favour, according to the file's verdict[skeptoid-640]: a storage jar.

And then a doubt at the very root of the story. A margin note in this file records that, according to later critics, König never saw the jar excavated[eggert-1996]. His find-spot comes from his own memoir, written after the fact. The context, meaning where exactly it lay and what lay beside it, is thin.

Even the date moved. König's label was Parthian; St John Simpson[bbc-simpson-2003] of the British Museum considers it Sasanian. A battery that cannot hold on to its own century does not inspire much confidence as a technological milestone.

Here is the fair other side, as the file states it. Supporters note that replicas reliably produce current[skeptoid-640], and argue that some purpose, such as mild electrical effects in a ritual, cannot strictly be ruled out. That is true in the way most negatives are true. But nothing in the record points to it, and the file's Finding reflects that: explained, with the popular claim marked debunked.

So where did the story bend? Not in 1938. König's title[konig-1938] ended in a question mark. The bend came afterwards, when the question mark fell off and the jar became the Baghdad battery, a fixture of books about lost ancient knowledge. The object never changed. The way people talked about it did.

There was, in principle, one way to go back and check. The jar itself. And that is where the file runs out of road.

The battery question is, for this archive, closed. The jar question is not.

In April 2003, as Baghdad fell, the National Museum of Iraq was looted, and the jar was reported lost[skeptoid-640]. Whatever residue it held, whatever marks the copper carried, whatever a modern lab could have read off the inside of that pot, may be gone with it. I find that the hardest line in the file. The one witness that could have answered our questions is missing.

Here is what the documents do not settle:

• Where exactly was it found, and in what context? The find-spot rests on König's own later account[konig-1940], read by us only through Eggert's citation[eggert-1996]. No excavation record is in this file.
• When was it made? Parthian per König, Sasanian per Simpson[bbc-simpson-2003]. The file leaves the dating unsettled.
• What did it hold? Scrolls, as Paszthory[paszthory-1985] proposed, or something else. The storage reading is the favoured one, but without the object or its residue it remains a reading.
• The original papers. König's 1938 article[konig-1938] is not online, and his 1940 memoir[konig-1940] has not been read in the original. Paszthory's page numbers and the BBC article date are also on our list to verify.

And here is what the file can say with confidence. A jar with copper and iron parts existed in the Iraq Museum. Replicas of it can produce a small voltage[eggert-1996]. That is the whole of what it shows.

What it does not prove is that anyone ever used it to make electricity, or that ancient Mesopotamia knew how to electroplate. The case against[skeptoid-640] stands on missing wires, missing plated objects, and a better explanation that fits a sealed vessel.

I don't want to end with a sneer at the people who loved this story. It is a good question, asked honestly in 1938, and replicas really do tingle. The fair hearing for the other side is in the file[skeptoid-640]: an unknown ritual use cannot be strictly excluded. It simply has nothing behind it.

The documents, the drawings and the timeline are below. Read König's question mark again before you read anyone's answer.