The Witch Bottle: Archaeology of a Counter-Charm

In 2004, workers from the Maritime Trust — a Greenwich-based charity that preserves historic sailing vessels — turned up a sealed stoneware jar about 1.5 metres below ground. It rattled. It splashed. An X-ray showed pins and nails jammed into the neck. The jar had been buried upside-down.

It was the first intact witch bottle to be opened and analysed scientifically in the United Kingdom. What was inside told a more precise story than anyone expected: 12 iron nails, one of them driven through a small leather heart; 8 brass pins; clumps of hair; 10 manicured fingernail clippings; a clot of what appeared to be navel fluff; and, confirmed by proton nuclear magnetic resonance spectroscopy, human urine. Traces of iron sulphide indicated that sulphur — brimstone — had also been included. The nail clippings were neatly trimmed, suggesting the maker was not a labourer.

The Greenwich bottle is the clearest single window into a practice documented across more than 250 English examples: the witch bottle, a sealed vessel of personal material buried at the threshold or hearth to intercept and reflect a witch’s curse.

The jar

The standard container from the late sixteenth into the seventeenth century is the Bartmann jug — also called a Bellarmine or greybeard — a salt-glazed stoneware vessel made in the Rhineland, primarily at Frechen, west of Cologne. The jugs are pot-bellied, about the size of a wine bottle, and moulded on the neck and shoulder with a grimacing bearded face mask that gives the ware its anthropomorphic quality: the bulging belly reads as a body, the face as a head.

The name Bellarmine attaches them to Cardinal Robert Bellarmine (1542–1621), the Jesuit theologian who was the scourge of the Protestant Netherlands. The association — that Protestants smashed or defiled the jugs by way of cursing his effigy — is plausible as a piece of popular culture but chronologically awkward: Bartmann ware was being produced and exported across northern Europe before Bellarmine rose to notoriety. The bearded face is more likely a generic apotropaic image, a leering guardian stoneware tradition that predates any particular target.

The jugs were workaday export goods, used as containers for wine and gin. Their ubiquity in England by the mid-seventeenth century made them the obvious choice of vessel: they were watertight, durable, they had a sealable mouth, and — not incidentally — their humanoid shape mapped neatly onto the sympathetic logic the bottles required.

Glass bottles appear in the record too, particularly from the later seventeenth century onward. Four complete Bartmann bottles were recovered from Abbots Ann, Hampshire in 1981 during British Telecom exchange work; they had been buried upright and inverted beneath a threshold, stoppered with their original corks. But the Bellarmine/Bartmann remains by far the most common form, accounting for roughly half of the approximately 250 English examples Hoggard has catalogued.

The contents and their logic

The contents of witch bottles are not random. Across the archaeological record, iron pins and nails appear in around 95 per cent of all identifiable contents. Urine is the next most common constituent, confirmed either by chemical analysis or by historical accounts. Hair, fingernail clippings, and fabric — often cut into a heart-shape and pierced with a pin — recur throughout the record.

The mechanism, as explained in contemporary sources, is sympathetic: the urine represents the witch’s own waterworks, and the bottle’s bulbous shape mirrors the witch’s bladder. Nail the pins and bent iron into the liquid, and you have reproduced the witch’s internal anatomy, under assault. Retired chemist Alan Massey, who analysed both the Greenwich bottle and the 2001 Felmersham, Bedfordshire example (found beneath a cottage hearth), summarised the operating theory plainly: “the nails and the bent pins would aggravate the witch when she passed water and torment her so badly that she would take the spell back off you.”

This is the key distinction: the witch bottle is counter-magic, not attack magic. The afflicted person is not cursing the witch unprovoked. The bottle is a trap — it captures the witch’s sent harm and redirects it. The witch bottle was understood by its makers as a legitimate defensive act, not an act of malice. This moral framing appears clearly in the seventeenth-century textual sources.

Joseph Blagrave’s Astrological Practice of Physick (1671) is one of the earliest printed descriptions of the procedure. Blagrave writes in the context of advising those who have suffered witchcraft — the bottle is a remedy, prescribed in the same breath as herbal treatments. The sympathetic logic requires that the materials come from the afflicted body: you cannot make a witch bottle for someone else using your own hair and urine. The bottle must carry the victim’s own personal matter.

Where they go

Hoggard’s database shows a clear pattern of placement: roughly half of recovered witch bottles were found focused around the hearth; approximately a quarter were beneath thresholds; the remainder turned up beneath floors, within walls, and in other liminal positions within the structure. Hearths and thresholds are not arbitrary choices. They are the two points at which the boundary of the house is most permeable — fire draws air down a chimney into the home, and the doorway is the literal opening through which visitors, animals, and anything else enter. Both are sites where a witch’s influence was understood to gain entry.

The deposit is permanent by design. Witch bottles were not dug up and refreshed; the logic of the charm required the bottle to remain hidden and intact. Hoggard records an early eighteenth-century case from Coopersale, Essex, where two glass bottles were sealed into a hearth surround in sequence — the second apparently added when the first was found to have cracked. The continued efficacy of the charm depended on the vessel remaining whole.

Who made them: cunning folk and self-help

Owen Davies’s work on cunning folk — the local practitioners of magic who served English communities from the medieval period through to the late nineteenth century — places the witch bottle squarely within their professional repertoire. When a household believed itself bewitched, consulting the local cunning man or woman was the standard first move, before (and sometimes instead of) going to a physician or magistrate. The cunning practitioner would diagnose bewitchment, identify the likely witch if possible, and prescribe a counter-measure. The witch bottle was among the most common prescriptions.

But not all witch bottles were professionally made. The Blagrave passage and similar contemporary sources read as instructions for self-assembly. The social spread of the practice is reflected in the archaeological record: bottles have been found in settings that range from modest rural cottages to properties of clear means. The Greenwich bottle, with its manicured fingernail clippings, suggests a maker of some social standing who had the leisure for a proper manicure. The Felmersham bottle, from a country cottage hearth in Bedfordshire, suggests otherwise. The practice crossed class lines.

Glanvill’s Saducismus Triumphatus (1681), a major seventeenth-century defence of the reality of witchcraft, records witch bottle use in case-study form, again framing it as a standard response to affliction rather than an eccentric folk survival. The practice had enough currency to be worth documenting in a serious theological polemic.

The record outside England

The witch bottle tradition spread to the American colonies with English emigration in the seventeenth century. Several examples have been identified in archaeological contexts in New England and the mid-Atlantic states. The American record is thinner — only a handful of confirmed examples against England’s 250-plus — which suggests the practice was exported but did not root as deeply, possibly because the specific social conditions of English cunning craft did not transfer wholesale to the colonial setting.

Within Britain, Scotland and Wales are notably underrepresented in the record relative to England. Whether this reflects a genuine regional difference in practice or a difference in excavation history and recording culture is an open question Hoggard’s ongoing data-collection is designed to help answer.

Making a witch bottle: the reconstructed procedure

The textual sources and archaeological evidence together give a coherent account of the procedure. In present-day practice, a reconstruction based on the historical record runs as follows.

Source a stoneware jar with a mouth narrow enough to seal — a Bellarmine reproduction is directly on-tradition, though any small-mouthed vessel with an opaque, durable body will serve. The historical containers were not decorative.

Cut a small heart from a piece of leather or red felt. Drive one or more pins through it. Add the remaining pins and iron nails — bent, as the historical materials usually were, which may indicate that bending the iron was part of the activation procedure rather than incidental.

Add personal material from the person to be protected: nail clippings, a lock of hair. If following the historical model closely, urine is the critical liquid component — its specific gravity, chemical signature, and biological connection to the individual are what the sympathetic mechanism requires. The historical record shows no substitute.

If including sulphur (brimstone), add a small amount. Iron sulphide traces in the Greenwich bottle suggest this was common enough to be a standard ingredient, though not universal.

Seal the vessel. Beeswax over a cork follows historical practice. Invert it — neck-down — and bury it at the threshold or hearth of the space to be protected, deep enough to be undisturbed. Mark the burial location in a way only you will recognise, if at all. The original makers clearly did not mark them: the bottles only surface when foundations are disturbed centuries later.

Leave it. The logic of the charm requires it to remain hidden and intact.

What to make of the practice

The witch bottle sits in unusually firm archaeological ground. Unlike most folk-magical beliefs, it survives as a physical object with a recoverable material record, a documented mechanism in contemporary sources, and — since 2009 — a scientifically analysed exemplar in the Greenwich bottle. This is not tradition reconstructed from folklore fragments or twentieth-century revival writing. It is a practice that can be followed as it was actually performed.

The objects themselves also carry something the textual record alone does not: the weight of long, deliberate concealment. The bottles were not displayed, not passed down as heirlooms, not mentioned in wills. They were buried and forgotten, and they have been found, mostly broken, across English building sites for three centuries. The Greenwich bottle survived sealed because it lay in ground undisturbed by later development. The rest of the record is fragments — which is, in its way, evidence of how widely the practice spread, and how completely it was meant to disappear into the walls of the house.

Cross-references on this site: (forthcoming — concealed shoes and dried cats as parallel apotropaic deposits; cunning folk and their professional practice; threshold magic across traditions.)