My head hurts

Conceptualising pain is quite a difficult thing to do, as it is a fundamentally subjective experience. Even in the 21st-century with the accumulation of medical knowledge, pain remains a regular or constant companion even for most people able to access the benefits of this modern medicine.

Analgesia (pain medication) provides a biomechanical pathway to resolve issues of pain, though it remains problematic because pain is incredibly diverse and is often an indicator of other biological issues which may then be masked through its reduction. Experiences of pain in humans are subjective but, at least to a point, allow for its essentialist consideration and comparison with our antecedents. The people of the ancient world felt pain and, at least mechanically, in the same ways that it is felt today.

I’ve been reading some more theoretical work on pain and am trying to think about it more deeply in terms of its potential to explain some ancient magical/medicinal practices. It is also designed to help me think about my own experiences of chronic muscle pains. The purpose of this brief blog is to highlight some of the varied magical treatments offered to the treatment of an undoubtedly common complaint in the ancient world – headache.  Partly because I have several headaches a week and I’m curious what could have happened to me in the ancient world.
Spells of the Papyri Grecae Magicae

PGM XVIIIa. 1-4. “”Lord Sabaoth, repel the pain from me, the headache pain, I pray, take [from me]. . . .”

A fragmentary text requiring what is probably a spoken prayer at the start, though if a material or gestural component were also required these are now lost.

PGM VII. 199-201. *For migraine headache: Take oil in your hands and utter the spell I “Zeus sowed a grape seed: it parts the soil; he does not sow it; it does not sprout.”

Allusions to the lack of growth of a seed and the reduction of pain(?). Spoken elements requires readily accessible material component as well.

PGM XCIV. 39-60. *Another, for migraine headache:

I. . . IYIO
. . . YAO~
. . . OYOO
. . . OY. El

“It delivers (?). And this ir Neros’ ‘ fmula: / “OURBEDEKAEIS OUROURBEDERAEIS OUROUROUBEDERAEIS EISTHES ABRASA ELECH BELLENOURE OUNOURE BAPHAMMBCH, to you I speak, pounding headache: don’t throb, don’t rage, / don’t shake the teeth, don’t produce mucus, don’t  produce a ‘black-out’, don’t stir up convulsions. For if there is throbbing, raging, shaking of teeth, producing of mucus, producing of a ‘black-out,’ I or stirrings of  a convulsion. . . sb . . . G . . .A. . . X . . . CH . . . /. . . A. . . .””

A much more complex spells requiring an amulet to be written and, presumably worn in the first instance. It is appended to by Neros’ formula – perhaps this is an additional spoken prayer invoking the voces magicae and various demonic names to enhance and improve the efficacy of the original amulet? It’s interesting that the migraine headache is being directly attended to, conceived as a figure that can be threatened in this way.


Marcellus of Bordeaux
Marcellus is a 4th/5th Century AD Gallic writer whose work De medicamentis records various contemporary medicinal cures.

“Magnetic stones, which emit blood and attract iron, tied to the neck or around the head, cure headaches” (1.63)
Is this a reference to haematite? In either case it present a straightforward solution to the problem – find the right material/substance and place it into contact with the area of pain. The magnetic qualities ascribed to it could have allusory relationships to the ideal of attracting the pain or, conversely, repelling it.

“Stone’s from a pig’s head, one of the right side and the other on the left, suspended on a thread is a cure for the head” (2.7).

This is similia similibus curentur through and through. Though where the thread is placed is unclear – on the body? In the house? Somewhere else entirely? Proximity is certainly a feature of many medical amulets and we may expect this to have been required to be on the body somewhere.


“Serpentine: Those that are black, ash-coloured, or with which lines are all useful when tied on to those who have been struck by vipers of who have headaches. “They say” that the one with the lines is particularly helpful for lethargic fever or headaches (5.153)”


A Magical gemstone

Recorded in the Campbell Bonner magical Gems Database, this gemstone in the Paul getty Museum is pale agate and inscribed on both faces in Greek. One side is a long palindrome and the other: “Deliver Gaia from the fever and also the chills and from her headache”.

This gemstone exemplifies some of the methods used in above magical practices – the use of arcane knowledge, exotic or specific materials, and requires it probably use by a specific individual. Note that Dioscorides mentioned fever and headaches together and that the two are also conceptually joined here.


Pliny the Elder

And finally: Not a cure but a preventative of sorts…

“As to the wines of Pompeii, they have arrived at their full perfection in ten years, after which they gain nothing by age: they are found also to be productive of headache, which often lasts so long as the sixth hour of the next day” (HN 14.8)

Peony Power

“Ephemerality is an often overlooked feature of Roman ritual practice. Particularly in ‘magical’ ritual practice, there are clear indications that efficacy can be influenced by intangible concepts. Thus, performing the right ritual, at the right time, in the right place, with the right materials was essential for the magic, in an efficacious sense, to work.”

So went the introduction to my presentation at TRAC 2019 titled ‘Flower Power: Plants, materiality and ritual efficacy in Roman magical practice’. The talk/paper/fevered ramblings intended to highlight the important of plants and their planty agency in magical ritual practice.

Publicly playing with peonies at TRAC.

There are numerous examples of plants as materia magica and/or materia medica (let’s not get bogged down in the semantics here) from the Roman world and wonderful instructions on how to use them. Thus we get Dioscorides suggesting that:

“Some people use the roots [of Monk’s Rhubarb] as an amulet against Scrofulous swellings of the glands, by tying them onto the neck”

And Marcus of Bordeaux to suggest a cure bleeding:

“|For bleeding] the root of a hemp plant tied to the right arm or suspended from the neck bound up with a cord”

Other ancient authors are, of course, available at all good digital libraries. In the presentation I dwelt briefly on one particular spell/instruction from the Greek Magical Papyri:

 “Wrap three peonies around your left arm and wear them” (PGM LXII 23-24).

The reason why is missing, but the instructions are there. During the talk I spun off into a quick phenomenological investigation on the variations of this particular spell; something I’d like to expand on here.

Peonies are plants. In modern biological terms there are about 33 species within the family Paeoniaceae, all located in the single genus Paeonia. Most are herbaceous perennial plants with compound, deeply lobed leaves and large, often fragrant flowers.

What I wanted to explore during the talk was that the instruction in the PGM leaves a lot of room for individual creativity in the ritual process and that there are numerous different results of these instructions. I argue that the exact position, composition, nature, colour, and the act of binding them were all individually meaningful and could have drawn on a wider body of magical or medicinal knowledge to enhance or instil this meaning.

The arm

The first point to introduce is the instruction to place the flowers on the left arm. In the Latin speaking world there was meaning to the word left: sinister as both ‘left’ and ‘unlucky’ or, perhaps more accurately, ‘inauspicious’ due its association with the auspicia (Nec coelum servare licet: tonat augure surdo, et laetae iurantur aves, bubone sinistro (Lucan. Pharsalia V, 395)). The left side was deliberately chosen, perhaps because it was the unlucky side, the side more open to supernatural danger. Pure speculation, but rather than this being a negative thing, perhaps the left side was deemed closest to the supernatural world because of this cultural relationship and thus it made sense to use that knowledge to harness part of it.

Tying to the arm also comes with a choice of where on the arm. During my initial stages of fumbling about I had constantly chosen the forearm, but when I asked my partner to have a go (having not seen the earlier attempts) she immediately tied them to the upper arm. Whether the flesh was exposed or covered by clothing could also have had implications for the ability to tie the flowers on: the sensory experience of having plants touching skin as well as mundane issues such as marking or staining cloth.

The plant

I have made an assumption that when the PGM listed ‘Peonies’ it meant ‘the flowers of the Peony plant’. Certainly we may question this interpretation as this comes down to contemporary knowledge – would a 3rd Century inhabitant of the Roman Empire know what a peony plant was if they had read (or been told) this instruction? I have absolutely no idea, how could I? – the fact that it existed in the PGM and had been translated suggests that the noun had meaning for someone and it would be incredibly dangerous to assume any level of ignorance within the ancient world based on my 21st century understandings of botany.

Leaves, roots, or woody plant stems may have featured but in my assumptions here we are talking about flowers. Immediately questions arise: The whole flower head? Individual petals? Flower head and stem? How much stem? All were potential points of creativity. Was the flower cut whilst in bloom, or immediately before or after? Were there specific tools to use to cut the plant? Did it have to be done in certain weather, or on certain days? At the very least, the biological clock of the plant will render a specific temporal window open to the user to find and use the plant.

Was the plant in a ‘living’ state or is it dried? A dried plant could be attached as a whole, or suspended within another container (textile bag for example) – certainly important for the sensory experience if not the ritual efficacy.

The binding

The third integral component is something to attach the plant to the arm with. Naturally the issues mentioned immediately above come into effect here. The first point to mention is that a separate binding was not necessary if the flower stem was long enough to wrap itself integrally around.

However, using a simple textile string or rope, flowers could be attached to an arm with relative ease, but therein lays another assumption. Could metal binding be used? Or other organics, like plant vines or roots? A sticky substance like honey or wax? Is there any need for a binding at all if clothing was being worn and it could be tucked or held beneath? All valid questions to my mind.

Returning to the assumed textile rope, there are opportunities here to build on other fragments of magical knowledge. Perhaps a ‘black of Isis’ (a piece of textile taken from a statue of Isis used in many other spells in the PGM) or the eponymous red wool could be used? The position of the binding may be relevant – the number of binds around the arm, their position, and indeed nature. A Hercules knot, for example, may be used to add an additional layer of magical knowledge to this already frothy ritual soup.

Wearing them

How long for? In what situations was this appropriate or not appropriate? Was this a visual cue to others, who shared the same ritual knowledge, that the user was unwell or trying to supernaturally protect their personal interests? Whether it was public or clandestine seems to be an important element in this, but again the space for ritual creativity allows for both.


Like at the conference, I’ve once again tied flowers to my arm to investigate some of these issues. I used five plastic peonies (as an analogue for real ones whilst travelling to Kent) and white shoelaces – hardly ancient materia magica in either case, but they served a purpose. Some thoughts on wrapping peonies around the left arm and wearing them:

  • Tying a binding to your own arm is not straightforward. It causes the hands and arms to comport themselves in an usual way. My fingers on the left hand were helpful only for holding a working end of the rope whilst the right hand did all the fine manipulation. Because of this, the upper arm was more difficult to bind to because of the loss of the left hand as an anchor point.thumbnail_IMG_20190426_083739
  • Having someone else do this to you makes it a lot easier!
  • Having failed several times to place the flowers first and then bind them, I opted for inverting this scenario and binding first and inserting flowers at positions of my choosing.
  • I used six wraps, crossing over each, and a Hercules knot. Using these materials, loops closer together were stronger and did not slip.
  • Tying it too tight, obviously created a tourniquet effect and was seriously uncomfortable.
  • Flower head towards hand or away from it?
  • Real flowers rather than plastic flowers would be easily damaged by this ritual action – petals could fall off; the stem could be bruised or cut by the binding. A recently cut and exposed stem would seep against the skin/clothing.
  • The proximity of real flowers heads to the body must have had a scent.
  • The left arm is subsequently limited in its effectiveness for, well, just about anything afterwards to prevent the delicate flowers being abraded or damaged by obstacles. Was there thus a minimum or maximum time to have worn the Peonies?



Clockwise from top left: One Peony slotted into pre-bound rope; Three arranged together at the wrist, though inversions were simple to do; Individual creativity – I used five.


Whilst there are no conclusions to be drawn from this rather daft experiment, it serves to highlight that even the most seemingly straightforward of Roman magical rituals could have been weighted by individual knowledge by practitioners and subsequently adapted or changed through the capacity for creativity. Plants may have afforded a strong sensory experience in a way that inorganic materials did not. We should reflect on both of these points.






Casing the joint: Amulet cases and the protection of materia magica

In May on 2017 I blogged here about an object in focus: a gold lamella designed for protection during childbirth. In this edition I want to revisit the practice, but not from the perspective of the epigraphic texts which contained magical formulae but from the containers which may be designed to carry them. This was inspired by a chance opportunity this week to handle one of the gold examples in the Yorkshire Museum during my day-to-today curatorial shenanigans there.

To introduce the practice: ‘Lamella’ does not describe a function, but a shape – a thin, metal plate (Betz 1992, 336). There is a rather grand and incredibly useful catalogue of them published by Kotansky (1994). These are all small and very, very thin. I am aware of 9 gold examples from Roman Britain and six very similar examples in lead alloy (probably a different blog there – we’ll focus on gold). Of the amulet cases I know of, from Roman Britain, 4 are gold, 1 possible example is lead, and there is 1 other possible candidate.


Figure 1: One of Nine – a gold lamella, handily also from York. This example bears two lines of Greek text invoking Chnoubis and was designed to treat stomach complaints. You can see it was also originally rolled. Associated with the Railway Station cemetery site. In the Yorkshire Museum. 

The recently discovered lamellae, reported through the Portable Antiquities Scheme, have been unrolled in order to read the text. The rolled lamellae are designed to be suspended in the so-called ‘amulet cases’ – hollow gold cylinders, sealed at both ends and suspended via two or three loops on one of the long edges. An example from Budapest, Hungary was found within one such gold, cylindrical capsule and another, from Krefeld-Gellup, Germany included a capsulate lamella associated with an inhumation burial (Kotansky 1994, no.21, fig. 23, and no. 4, fig. 5 respectively). The gold capsule case of a lamella from Vienna, Austria (Kotanksy 1994, no. 17), was also associated with an inhumation (in a sarcophagus) but contained an inner capsule of silver, itself containing another capsule of bronze which held a lamella in the innermost layer.

An amulet case is not, however, always an indicator that a lamella was originally present within it. Some examples did contain not a lamella at all but a powdery substance of presumed medical or magical relevance (at least three of the examples in the British Museum are reported to be filled with sulphur: BM: 1917, 0601.2984; BM: 1917, 0601.2981; BM: 1981, 0201.30, the latter from the Thetford Treasure, Norfolk).

Is this the same practice? Visually it would certainly have looked identical to an onlooker or, even, the user – a golden tube suspended on or about the body somewhere. Were the contents known to the individual? I think we have to presume that this is the case, especially given the bespoke (and sometimes) time-limited nature of the lamella text; think here of the lamella mentioned at the start, designed to be used only during children for a specifically named individual.

Unless either type, the rolled metal sheet or the powdered substance, were tightly packed into the tube they would be both liable to movement within the tube. Of the two, the sheet can be carefully folded or rolled to size for a better fit: the lamella from Sagvar (Hungary) was folded only once on its long axis and close to one side, suggesting it needed narrowing only slightly before it could be placed into a tube (Kotanksy 1994, 83).

Figure 2: The gold amulet case from the Church Street Sewer, York, 1972. Note the two suspension loops, single closed end, and single open (broken) end. In the Yorkshire Museum.

Looking more closely at the York example it does not exhibit any clear evidence for long-term use-wear on the suspension loops –  an idea that, perhaps, chimes closely with the need for temporarily highlighted above. Whilst I am now unable to account for the taphonomic processes that this case has undergone, it does retain one of its closing end-pieces entirely in situ and the other is compressed into the tube itself. Unless this was undertaken in post-excavation (unlikely?) I can only accept that this is the condition in which it was discovered, in the Church Street Sewer in York in 1972. Now, the exposed end almost exactly bisects the diameter of the tub, making the possible removal of a rolled lamella more difficult unless it was a very poor fit for the case itself and/or a sheet was removed and the cap pushed back in.

Powdered materials in cylindrical cases could be tightly packed though it is unlikely that they could be completely compressed to prevent any sort of movement of the material. Perhaps the movement of the material was of relevance? The weight distribution of the pendant filled with powder may change from side to side. This may be an indication of why the suspension loops for the cylindrical cases are at either end rather than just in the middle – to prevent the tube from tipping over to one side. As far as the York example is concerned if it contained a powdery material this could very easily has entirely escaped the capsule if the end-piece was damaged in the way that it is.

In either instance there is an important question here – was it deliberate? It’s discovery in a sewer system, fed from both the Praetorium bath-house and the legionary’s bath-house, is easily explained away as an accidental loss, but this only masks the more important question of what someone was doing wearing a gold amulet case in a bath-house. The difference between casually losing a golden amulet and taking it apart and throwing it down a drain is pretty huge, especially if it was a thing that is fundamentally designed to improve the lot of an individual. Perhaps there was a situation in which deliberately destroying an amulet was an appropriate thing to do? How this related to its disposal in a bath-house we can only really speculate. Perhaps it’s contents had fulfilled their purpose, or their time-limited duration had run out? Perhaps it hadn’t worked at all? Perhaps the contents were designed to be kept until required (“break in case of emergency”?) and the whole was worn on or around the time when whatever ailment or problem was envisaged to rear it’s ugly head.

I’ve made an assumption here, that it is being worn (it could have easily been disposed once damaged elsewhere), but making this assumption leads to the suggestion that it was being worn in such a way that is conducive to being, otherwise, naked – i.e. somehow physically suspended from the body, via a chain or cord on the neck, arm, or tied into hair.

The gold is itself not a cheap thing to come by, requiring financial access to this precious material as well as a gold-worker to create the case. In this sense it might not an especially egalitarian practice, but if we can associate it with the well-paid legionary garrison or a member of the Legate’s household then we can at least accept that these groups are more likely to have access to this sort of material (especially the legate and his family). Interesting that the cases were largely unadorned with any additional decoration or inscription. One might imagine that apotropaic marking or symbols could be added to enhance its efficacy, but un-dramatic, smooth-surfaced gold seems to have been the required standard for these tubular cases.

One of the most intriguing questions I find here is – Is the amulet case purely designed to protect its content or did it serve an apotropaic function as well? It’s a question that I could certainly expend a lot of column-inches on, but suffice to say that there is a clear material link between gold objects within a gold contains or, a pale yellow powder in a gold-container. This colour link is contextual but, I would certainly argue, quite integral to their function. Thinking of the wider sensorium of the object, the colour is bright and simple and the gold would not tarnish and easily catch the light if worn over clothing. In the shadowy half-light of some of the rooms in a bath-house, it may have been particularly eye-catching. The amulet case is also incredibly light and even with the supposed contents mentioned above it would weigh no more than a few grams. That said, suspension loops could allow for suspension alongside other objects (glass beads spring to mind, but other apotropaic objects could also have worked). Perhaps they could be sewn onto textiles, or carried in a larger container?



Betz, H. D. (ed) 1992. The Greek Magical Papyri in Translation (Including the Demotic Spells). 2nd Edition. Chicago and London, University of Chicago Press.

Kotansky, R. 1994. Greek Magical Amulets: The Inscribed Gold, Silver, Copper and Bronze Lamellae. Part I: Published Texts of Known Provenance (Papyrologica Coloniensia XXII/1). Opladen, Westdeutsche Verlag


Testing ABRACADABRA: Sammonicus’ Magical Amulet

The late 2nd– early 3rd Century Roman author Quintus Serenus Sammonicus is particularly notable amongst researchers of ancient magic and medicine for being the first to propose the use of the word ABRACADABRA, and in treatment of tertian fever. The traditional ubiquity of this word in association with modern magic tricks has highlighted this use, though it would not have seemed unusual as a Roman phylactery.

The instructive text reads:

“Rather more deadly is what in Greek words is commonly called hemitritaion. This no one could express in our language, I believe, and neither did parents wish for that. Write on a sheet (of papyrus) the word ABRACADABRA, repeat it rather more often underneath, but omit the last letter, so that more and more individual letters will be missing from the lines, the elements that you remove, which you continually snatch away, while you commit to writing the others, until a single letter is to be rendered as the narrow end of a cone. Remember to attach this to the neck with a linen thread.”

There are already two excellent blogs about it here and here, considering it from different perspectives. I’m here to add a third.

In the interests of research, blog writing, and Friday afternoons I have opted for a spot of experimental archaeology and I will re-create Sammonicus’ charm.

Using papyrus (yep I have some hanging around for just this purpose) and a black ink pen, I drew the ABRACADABRA word series on two sizes of rectangular cut papyrus. The, arguably, small sizes chosen is informed mainly by the imminent need to attach it to my neck and the efficient use of the papyrus. There is no reason that this couldn’t be much larger, though doing so would change how I might conceive the attachment.


Figure 1: Abracadabra! Two magical papyri appear before your very eyes. 

Thoughts on writing

  • I am writing this on modern replica papyrus, so a decent analogue although I’ve never been able to compare with original papyrus as to its tensile strength and ink-taking properties
  • I used a modern, graphic design pen with black ink and a fine nib so writing the text was easy. Probably far too easy, it would be more difficult to do so with a stylus. Perhaps greater familiarity with a stylus and ink would also render using those tools easier, so I am being bit essentialist with this thought. Bad.
  • I was not particularly careful with my lettering and opted for haste; it took less than a minute to write the text.
  • Familiarity with the Latin alphabet is helpful. Being illiterate might increase the complexity of this somewhat, but the shapes are not impossible to copy without having had prior knowledge.
  • Familiarity with the word (again – modern usage) is helpful.
  • It could be written just as easily in Greek letters.

Thoughts on the text

It is not a palindrome if read conventionally, but the diminishing triangular format produces a repetition of the word written on the right hand side if read using the exterior letters from the apicial A upwards and to the right. This creates a palindrome of ABRACADABRARBADACARBA on the upper and right sides of the triangle. The left side is a repeating vowel series of the letter A, which is also thus at the start, end and centre point of the palindrome. Both palindromes and vowel series are frequently components of Roman magical literary practice. So too is triplication (three sides).

This is also a very simple spell in the grand scheme of things. Simple, repeatable, memorable even, and easily communicated verbally.


There is certainly a lot of scope in the instructions for individual creativity in how to wear this thing. Perhaps one major consideration is whether the next is folded or rolled up in any way. Both are certainly possible, as is leaving it open. I assumed that the text would be rolled, as this is the case with curse tablets and gold phylacteries. The latter were sometimes inserted into a gold or bronze tube (a.k.a amulet case) to protect the text and/or supernaturally enhance it. A textile pouch could also easily cover and protect the contents in such a way to allow them to be hung around the neck. Given that this text is made from papyrus as is designed to treat Tertian fever (i.e. it lasts about 36 hours) it, perhaps, was not designed for durability in mind and damaging exposure to the elements may be minimal.

In this scenario, I consider the charm to be without cover, leaving only the issue of affixing. Without too much effort, I came up with at least five different ways of attaching the papyrus to a linen thread (note that I used faux leather cord): by threading through cuts in the sheet; by using a hitch; by wrapping the thread around the sheet; by putting the thread through the centre of a rolled sheet; by knotting the thread to the sheet. Beyond these methods, the papyrus could also be wrapped or tied directly against the skin. Inevitably, other options were available and many of these could be used in combination.


 Figure 2: Some possible methods of attaching a papyrus sheet to a thread.

The use of knotting or wrapping could be particularly important as a binding gesture. I experimented only with the larger sheet, but all methods would be applicable to the smaller sheet as well, indeed it could be more effectively hidden if required.

Sammonicus was not specific on where about the neck to attach it. Clearly it could be worn as a pendant (around the neck and against the chest), or much closer as a choker; the position of the text may be important but this can be moved around the body. The former scenario affords an opportunity to hide the charm beneath clothing if so desired.

 Figure 3: Hung loose at the chest, tight against the neck, and loose in a textile bag. Again, many other variants are possible. 


Sensory experience

The creation of such a charm requires sufficient light to write the text, and the cutting, ripping, rolling, or folding of the finished document to facilitate its attachment. Such an object could certainly be purchased ready-made, or pre-made and stored in advance of illness.

It is lightweight against the body in this form, though would be heavier if encased in a metal tube. Generally, it is unobtrusive, though I can imagine it might be uncomfortable in the throes of a raging fever.

I didn’t test it specifically, but it seems apparent that bodily fluids, particularly sweat (from the fever), could penetrate an un-covered charm like this to its physical detriment. The sweat might adhere the papyrus to the body, perhaps intentionally?

Final thoughts

I strongly suspect that the lack of clear instructions on how to use the charm was not particularly important, as the all-important arcane knowledge of this practice was based on the use of the mystical word written in a non-normative way. Individual creativity could have played a key role in this and afforded an opportunity for personal choice to influence the space and place in which it was utilised.

Equally, it could easily be combined with other magical or medicinal practices and/or rituals. Like I mentioned above, this one is pretty straightforward, easy to remember, and easily communicated. It is certainly unclear whether this charm is designed to prevent the onset of fever, cure it once it has arrived, or a combination of the two – again, perhaps the ambiguity is important as it was up the individual practitioner?



Painkillers? Pierced teeth pendants as amulets

Museums and excavation reports from Roman Britain often contain a reference to a pierced animal tooth, perhaps used as an amulet. The tooth is usually a dog’s canine, though wolf canine, sheep incisor, bear canine, and pig/boar tusk also feature. What is abundantly unclear, however, is what these things were supposed to be used for.


Fig.1: A pierced canine canine (ha!) on display at Verulamium Museum.

Pierced dog’s teeth are referenced by Pliny (HN 11.63) as a ‘valuable amulet’, particularly efficacious for use with curing ‘childish terrors’ and as a medicine for teething (HN 28.78). He also mentions other possible uses for the teeth of horses (ibid.) and crocodiles (HN 28.28), the former performing similar functions to the pierced dog’s teeth and the latter as an aphrodisiac or as part of a ritual cure for fever.

It is one reference, but let’s run with it for the purposes of academic curiosity to see where it leads.

At the outset I should point out that other materia medica had been suggested for this purpose. For example, Dioscorides (De materia medica 1.71) suggested that

“teething and the itching that this causes are relieved if you smear butter on the child’s gums, either by itself or mixed with honey; The broiled brains of a hare, rubbed on and then swallowed, are equally effective. Eating a boiled mouse also helps…”

Both ideas appear in the 3rd century work of Serenus Sammonicus (Liber medicinalis 58.1029):

“A cure for teething: nature sends mankind in to the world naked at first. She then afflicts us with torture when she arms us with snow-white teeth. So bind round your child’s soft neck a horse’s teeth, the first teeth that fall out as a foal grows bigger. Or smear the child’s tender gums with the brains of a pig or of a hare, or with the snow-white milk of shaggy she-goats”

Looking into this, let’s start with some essentialism: teething hurts and there is very little that can be done to alleviate it. The NHS provides the following symptoms:

  • your baby’s gum is sore and red where the tooth is coming through
  • one cheek is flushed
  • your baby is dribbling more than usual
  • they are gnawing and chewing on things a lot
  • they are more fretful than usual

Teething is a variable process: some babies are born with their first teeth, otherwise the process may begin at 4 to 12 months of age and it may be entirely painless or partly/thoroughly painful. The early imperial writer Soranus mentions the occurrence of teething at about the 7th month (Gyn. 2.49) and suggests that massage of the gums may help, also that the use of a piece of animal fat from which the infant may suck the moisture is a good idea. Between Soranus, Dioscorides, and Pliny then we have two major modus operadi for alleviating the pain of teething: something soft and moist placed in the mouth as a lubricant, or the use of an amulet. Pliny isn’t revealing on where the pierced teeth should be placed, but that issue raises further interesting questions.

Let’s start with the animals – dogs & sheep primarily. Well, dogs are easily linked to medical practice through being particularly associated with Asclepius and considered to be possessed of healing qualities (Jackson 1988, 142). Sheep are more difficult, but the teeth we find pierced are the front incisors. No coincidence perhaps, that it is the incisors that are the first teeth to appear in humans? I’ll ignore the bear and pig/boar for this blog as they are differently sized pendants attached in different ways, and focus on the dog and sheep teeth because they are very similarly sized and, when pierced, almost always singly and through the root rather than the cap. How they work is certainly a different matter; we can infer a level of sympathetic magic – like cures like. Perhaps the adult tooth, disassociated from its original owner, was designed to focus the pain within itself, or to encourage the child’s teeth somehow. Within only a couple of lines I can only speculate, but suffice to say that within the panoply of Roman magical and medicinal materials, it doesn’t seem out of place.

Looking briefly at where these things come from in Roman Britain, it is clear they turn up in ones or twos from single sites, mainly the urban settlements. Of the 82 so far gathered as part of my PhD research, 23 come from stratified contexts ranging from the mid-1st to the later 4th centuries AD, with a tentative skew towards the first two centuries AD.


Fig.1: An impenetrable heatmap showing the distribution of pierced teeth pendants in Roman Britain: they’re pretty much everywhere. 

Accepting that pierced teeth pendants are potentially used for the treatment of problems with teething, we must question the point at which such an object was employed. Were they used in the weeks or months before it was presumed to begin (the application of prior knowledge by the biological mother, her own mother, or a midwife may be utilised here)? From birth? Only once the teeth start showing? Only once pain began? Perhaps of most importance is the question of ‘who is wearing the amulet?’, child or adult?

Whether a small infant would ever wear a necklace is open for discussion. Infants begin to place hands and nearby objects into their mouths during this ‘transitional feeding stage’ at about 6 months of age (Arvedson 2006, Table 3) so a small tooth nearby could present a choking hazard, unless it or the child was somehow immobilised. The young toddlers depicted on the Ara Pacis are shown to be wearing bullae, which are much larger than the pierced teeth pendants and worn at an age where it may be more appropriate to do so. Newborns may have been swaddled; there is both literary and visual evidence to suggest this may have occurred for about the first 40-60 days (Soranus, Gyn. 2.14; Croom 2000, 117-118). This is most likely to have occurred before the child had teeth, but would have allowed for a small amulet to be attached easily to a child. Bullae strung on long suspension straps and other chains of pendants are depicted on ceramic votives of swaddled infants from Paestum in the final three centuries BC (Ammerman 2007, 142f, figs. 7.12-14, 7.16-17).

Teething is closely associated with weaning, though the two processes are not necessarily concurrent. Soranus (Gyn. 2.47) suggests that breastfeeding continues, with supplements of solid food, up to the age of 18-24 months. An isotopic study of infant feeding practices in Roman London supports the idea that breastfeeding occurs at the very early stages in the life of the infants sampled and also that it continued to form part of the diet during a very gradual weaning process lasting, in some cases, up to the age of three years (Powell et al. 2014, 101).

The dual process of weaning and teething may in fact result in pain for both the infant and the adult if she remained breastfeeding during this time. It is worth considering that the tooth may, in fact, have been worn by the adult during this period – for example, if on a necklace and worn at the chest it could be encompassed between mother and child during the process of breastfeeding. Could both wear one? Perhaps a matching set from a single animal – left and right together? Or different animals for adult and child?

I’ve raised the question of temporal use here – was there a specific time where it was more, or less appropriate to use such a pendant? Well, presumably the efficacy is limited to the time that the child is teething (unless the teeth were possessed of alternative sympathetic properties), but was it worn only when the pain flared up or all the time? Likewise, if the adult wears it, is this likewise subjected to specific times?

Beyond the immediate use for one child, within a family was a new tooth required for each new child or could they be recycled? If the latter, might such teeth have been curated and used for multiple children, multiple families and even multiple generations? Did the family have to acquire the teeth from the animal itself or should we presume a trade in these animal parts?

Clearly many questions about the use of such a pendant can be posed and we can only speculate at the point of deployment for an appropriate apotropaic pendant. The fact that the teething process may be entirely painless could establish links between a pendant and its efficacy in stopping pain. Re-using such pendants for additional infants (assuming there is not a destructive element in the life-course of the object, i.e. it is destroyed once teething is over – there is no clear evidence for this), this may create a familial or interpersonal link between the object and the individuals involved establishing a shared memory experience between the object and its action (see Graham 2011, 25ff). In such circumstances the retention of the tooth pendant may not be surprising. Teething is, potentially, a very sensory experience; an upset, screaming, red-faced, dribbling infant is undoubtedly a worrying prospect for a parent and the opportunity to attempt to resolve such problems no doubt would be pursued.

I’ve raised many unanswered questions here, but I will be discussing them in greater detail with my paper at the MagiCog: Cognitive Approaches to Ancient Magic Conference at the ICS, London in January 2019.



Ammerman, R. M. 2007. ‘Children at Risk: Votive Terracottas and the Welfare of Infants at Paestum’, Hesperia Supplements 41 (Constructions of Childhood in Ancient Greece and Italy). The American School of Classical Studies at Athens. 131-151.

Arvedson, J. C. 2006. ‘Swallowing and feeding in infants and young children’, GI Motility Online [doi:10.1038/gimo17] (accessed 22/07/2016)

Croom, A. T. 2000. Roman Clothing and Fashion. Stroud, Tempus.

Graham, E-J. 2011. ‘Memory and materiality: re-embodying the Roman funeral’, in Hope, V. M. and Huskinson, J. (eds) Memory and Mourning: Studies on Roman Death. Oxford, Oxbow. 23-39.

Jackson, R. 1988. Doctors and Diseases in the Roman Empire. London, British Museum.

Powell, L. A., Redfern, R., Millard, A. R., and Gröcke, D. R. 2014. ‘Infant feeding practices in Roman London: evidence from isotopic analyses’, in Carroll, M. and Graham, E-J. (eds) Infant Health and Death in Roman Italy and Beyond (JRA Supplementary Series 96). Portsmouth, Rhode Island, Journal of Roman Archaeology. 89-110.


Object in focus: The love-charm of Polemius

A 1935 discovery during early excavations at the northern Roman town of Corbridge, Northumberland yielded a singularly fascinating magical object for Roman Britain. The object in focus here is a gold finger ring, decorated in openwork and consisting of sixteen facets, the edges of each are decorated with openwork peltas and each facet contains a letter or a leaf-stop.


The effect is to create elaborate see-through lettering. The letters are all Greek and spell out the phrase:

ΠOΛEMIOYΦIΛTPON (transliteration: Polemiou philtron) which translates as “The love charm of Polemius”.

In a recent blog I discussed the importance of written words and phrases to install good luck on an object or the person who uses it and this ring appears to use the same sort of magical mechanism – words as a source of supernatural power (van den Hoek et al. 2015). It was found in the relatively metropolitan site of Corbridge so the use of Greek shouldn’t come as a shock as the Greek language did feature throughout most of the province of Roman Britain. The ring is in the British Museum and published in Roman Inscriptions of Britain II (Frere and Tomlin 1991, no. 2422.12).

There are a great many betrothal charms from Roman Britain which are worn as declarations of love, often showing two people kissing or right-hands clasped in the dextrarum iuncto – a symbol of marriage. Other love tokens use phrases such as ‘My Darling’, ‘Sweet love’, or ‘Live for me’ and thus the Polemius charm doesn’t really fit into this category of ‘Romantic love’.

It may be presumed that the ring was owned by a man named Polemius, but it is possible that the name relates to the craftsperson producing the object. It is also, somewhat unclear how such an object may be used. Following the presumption that it is owned by Polemius, he may have worn it himself as an object intending to provide support in his love life (love life/sex life, who knows?) – in the same way that a Utere felix device might promote good luck and a generalised supernatural benefit to its user, this ring aimed to use the power of words to achieve this more specific function.

Perhaps he may have intended to give the ring to the object of his affections? Whether this was with or without their knowledge could be of serious interest. As a declaration of love between two people, it makes sense as an affectionate love-token, but as a clandestine object secreted on or around a person to influence their emotions through magical means it takes on a wholly different meaning. The specific ritual use of a ring in this manner is certainly unclear – the Papyri Grecae Magicae and Papyri Demoticae Magicae offer no real clues because, despite the existence of a number of spells and rituals which utilise a finger ring (PGM VII.628-42, XII.201-69, XII.270-350; PDM XII 6-20) all require the use of a setting or intaglio as the focal element of the magical rite rather than an inscription incorporated into the band itself. Thus we might consider the ring a bespoke creation – a unique thing made for or by Polemius.

Many of the more familiar Greek Magical gemstones (see the Campbell-Bonner database of the international corpus  and an earlier blog on one such example from Britain) incorporate Greek lettering or whole Greek words into their imagery. Polemius’ charm does not, however, use many of the other ‘magical’ elements (voces magicae, vowel series, palindromes, images or names of Gods, demons and other supernatural entities) visible in such examples.

The singular named link, of Polemius (and I am assuming that this name refers to a real human rather than a deity, demon or other supernatural force) intricately ties the efficacy of this object to one man. The question is whether Polemius was the person who created other charms on demand, as ritual practitioner (dare I say it, ‘magician’?) and it was his esoteric knowledge that formed the functioning part of the ring, or whether it was the name of the person who used it in one of the ways I briefly describe above.

Other openwork rings made from gold are personalised with names, though with less specific intensions.  In this category we can include those promising long life to Aemelia (Frere and Tomlin 1991, no. 2422.1), also from Corbridge, Eusebius (Frere and Tomlin 1991, no. 2422.5) from Bedford, and Olympios from Stonham Aspal, Suffolk (Frere and Tomlin 1991, no. 2422.10). Those of Aemelia and Eusebius are written in Latin, Olympios’ in Greek.

Given the explicit named link to one person and the frequently efficacious links in magical epigraphy between objects and individuals the ring clearly was designed to fulfil a very specific problem. We might assume that Polemius was having particular problems in the love department and this ring was an attempted solution. Helping Polemius to fall in love, or for someone to love him, or for a relationship to be retained etc. The ring could have served a medicinal effect instead of a magically emotional one – as a supernatural solution to a physical problem. Impotence perhaps, or infertility?



Frere, S. S. and Tomlin, R. S. O. (eds). 1992. The Roman Inscriptions of Britain, Volume II, Fascicule 4. Stroud, Administrators of the Haverfield Bequest.

Van den Hoek, A., Feissel, D. and Hermann Jr., J. J. 2015. ‘More Lucky Wearers: The Magic of Portable Inscriptions’, in Boschung, D. and Bremmer, J. N. (eds) 2015. The Materiality of Magic (Morphomata 20). Paderborn, Wilhelm Fink. 309-356.

The Walking Dead?

The twitter account of the Museum of London Archaeology (@MOLArchaeology) posted a link on June 8 to an interesting burial practice from Roman London which features in the Museum’s current high-profile exhibition ‘Roman Dead’: the use of iron shackles irreversibly attached to the limbs of the deceased – an idea, promoted in their blog on the subject from March 2016, which may be linked to a fear of the dead by the living.

As Mike Marshall pointed out on twitter, and MOLA themselves suggest in their blog, the term “shackles” is somewhat misleading as a shackle is a chained and removable form of bondage – these iron rings were neither. @MOLArchaeology continued: “The iron ring from the Crossrail excavations at Liverpool St. could not be opened and was probably too small to slip off, suggesting it had been forged onto the arm shortly before or after death”. And expanded the corpus with “Two burials with leg shackles were found at Finsbury Circus”, then postulating the interesting question whether this was “…a ritual to brand the bearer as a criminal, prisoner or slave in the afterlife or deny them final rest? Or did these Romans fear that the dead might walk again?”. The thread finished with a link to a similar practice from the Driffield Terrace excavations in York.

Well what a question – did the use of non-removable iron bands represent a material manifestation of a fear of the dead or the un-dead? Well for starters we should point out that the iron rings often accompany the mutilation of the skeleton somehow, like by perimortem beheading and this is a relatively unusual thing to come across in Romano-British burials. This topic is helpfully featured also in an article in the publication of the 2015 Theoretical Roman Archaeology Conference “Fear of the Dead” (Quercia and Cazzulo 2016) wherein the authors identify both these and additional strategies for the restraint of the deceased from northern Italy: straps which bound the arms and legs; nails in the burial and, in one example, driven into a skull; the removal of the torso and inclusion of an unguentarium in the pelvis; removal of feet; and the use of a wooden board strapped around the neck.

These examples are, to put it mildly, unusual in the grand scheme of Roman burial practices but a fear of the dead is not an unreasonable explanation for these practices. There are numerous textual sources in the Greek Magical Papyri (PGM, see Betz 1992) which advocate the use of the dead as a source of power or as a supernatural medium. A passage in PGM IV.2145-2240 illuminates the effect of such magical practices:

For wrecking chariots: Burn garlic and a snake’s slough as an offering, and write on a tin plaque “NEBOUTOSOUALETH BEU ERBETH PAKERBETH and ONOUPH; Overturn him, NN, and his companions”. Bury the tablet for 3 days in the grave of someone who died untimely; he will come to life for as long as it stays there”.

A passage in PGM XIII. 1-343 goes a step further and attempt to fully resurrect the dead:

Resurrection of a dead body: “I conjure you, spirit coming in air, enter, inspire, empower, resurrect by the power of the eternal god, this body; and let it walk about in this place, for I am he who acts with the power of Thayth, the holy god”.

If the iron work and mutilation were attempts at preventing the dead from rising up again, what prompted the decision to add the iron rings or remove a head? The untimely dead, featured above, the ‘Biaiothanatoi’ (Quercia and Cazzulo 2016, 30) may have been particularly vulnerable to the effects of magical practices. Premature passing, suicide, death in childbirth, murder victims, and others were the ‘restless dead’ – those who went before their time. Their exploitation was not necessarily for the purpose of recreating a zombie apocalypse, they may have been used as vessels to deliver a message or direction to a supernatural entity or, in some conceptual way, their un-used potential, power, ‘essence’, ‘soul’ etc. could be harnessed by the living. PGM IV.296-466:

Wonderous spell for binding a lover: Take wax from a potter’s wheel and make two figure, a male and a female. […….] And take a lead tablet and write the same spell and recite it. And tie the lead leaf to the figures with thread from the loom after making 365 knots while saying as you have learned, “ABRASAX, hold her fast”. You place it, as the sun is setting, beside the graves of one who has died untimely, placing it beside also the seasonal flowers”.

This link to the untimely dead does not, however, fully answer the question as such deaths must have been far more commonly encountered than the burial practices under discussion here. It is, at least, a stepping off point for trying to make sense of the thing. The chap from the Crossrail excavations was an adult, but had not died of old age and nor had any of the examples restrained in York. It’s certainly an issue studying Roman magic that it gives you a tendency to see magic everywhere, but these burial practices are good candidates for representing material strategies to, perhaps, prevent the deceased from helping anyone. Could the deceased rise from the grave of their own accord without living human agency involved? Again the weight of the numbers of untimely dead devoid of such interventions suggests otherwise or, at the very least, suggests that a different strategy was used that did not leave an archaeological trace.

Quercia and Cazzulo (2016) also suggest that particularly deep burials and those capped by stones could also represents attempts at some sort of restraining of the dead as things that were fundamentally unmovable by an individual. These may also have been things that have been overlooked throughout the past centuries of archaeological enquiry! Iron rings, in and of themselves, were probably uncomfortable but may not have completely prevented physical movement by a living person wearing such a thing. Perhaps then they were an allusion to restriction? Did the material matter – why iron? Good ol’ PLiny (HN XXXIV) at least suggests various medicinal uses of iron and iron rust. Presumably there was a logic in being able to forge straight onto a limb that meant iron was the only real choice. On a living participant that would also have been an intensely painful procedure, but is there more to it.

It’s certainly a bit of a leap of faith, but why do we not also consider the ‘deviant’ ringed and mutilated burials as attempts to prevent something being done to the remains rather than the remains themselves being the pejorative agents in this? Perhaps that’s just my retrojected assumptions about how the un-dead ‘work’ from modern fictions and folklore surrounding it? I’m not sure if anyone else has made the claim yet, but the practices of binding and restraining of the dead in such a manner could have been a realistic attempt at protection of the deceased themselves from the living rather than the other way around. Believing that practices which exploited chthonic powers through the recently dead were undertaken could have prompted the pre-emptive projection of the dead by preventing them from walking around in thrall to any living person with a spell book and the correct materia magica.

A reference to the use of iron rings, touched on in PGM IV.2125-39, offers another suggestion:

“A restraining seal for skulls that are not satisfactory [for use in divination], and also to prevent [them] from speaking or doing anything whatever of this [sort]: Seal the mouth of the skull with dirt from the doors of [a temple] of Osiris and from a mound [covering] graves. Taking iron from a leg fetter, work it cold and make a ring on which have a headless lion engraved. Let him have, instead of his head, a crown of Isis, and let him trample with his feet a skeleton (the right foot should trample the skull of the skeleton)…”

Immediately from this example there are allusions to the practices of using an iron ring (note: large enough to fit on a human(?) head) and the use of feet, but the additions of these materials are designed to restrain or seal the remains from being useful to anyone else. The important thing to take away from this example is that the additions of these materials may have been by someone attempting a magical ritual themselves – accessing a grave shortly after internment may have been difficult, but certainly was not impossible. Rather than a caring family protecting the deceased, such examples may represent the attempt at silencing or sealing the dead – to restrain them in their graves in an attempt to protect the living from their knowledge or interference rather than their undeadness. In this scenario the ‘living’ means an individual/group with a vested interested in the dead staying that way.  Thus in this capacity it may be that the dead weren’t being restrained because it was the fact of their coming back to life that was the main concern; not only was this believed to be possible, but they may return with information or knowledge that was detrimental to some of the living. Any sane person would do the same, right?

Pure speculation of course, all of it, but my goal with these musings has been to take the conversation towards understanding the behaviour in a more nuanced way and accept that these practices were reacting to quite specific social situations in specific places rather than a generalised fear of the dead rising up again.



Betz, H. D. (ed). 1992. The Greek Magical Papyri in Translation.

Quercia, A. and Cazzulo, M. 2016. ‘Fear of the Dead? ‘Deviant’ Burials in Roman Northern Italy’, in Mandich, M. J. et al. (eds) TRAC 2015: Proceedings of the twenty-Fifth Annual Theoretical Roman Archaeology Conference. Leicester, TRAC. 28-42.



Jet Bears and Magical Stars

As part of my PhD research I have been travelling around the UK to check out museum collections; part of this involves coming across objects I had read about but not seen in the real world. Perhaps the example of this category which brings me most pleasure is the disparate group of tiny jet figurines which depict a bear. Seven such bears are known to modern archaeology from Britain: Four from Colchester, two from York, and one from Malton. A further two are known from Cologne and Trier in Germany. The individual finds are well published in Nine Crummy’s 2010 Britannia paper ‘Bears and Coins: The Iconography of Protection in Late Roman Infant Burials’. As the title suggests, there is a strong correlation between these scant finds and infant burials in the late Roman period. In addition to the bears there are two functionally and morphologically similar Jet ‘big cats’ – from Chew Park Villa and Chelmsford.

FIGURE 1: Jet bears from Malton (left) and Bootham, York (right). Respectively in the collections of Malton Museum and Weston Park Museum, Sheffield. Low-res images by the author. 

The distribution of these finds in York, Colchester (and Eastern Britain generally), and also Cologne is paralleled by the 22 known jet gorgoneia pendants (12 from Britain); these have a strong association with inhumation burials of young-adult females in the 3rd and 4th centuries AD (Eckardt 2014, 112-116; Parker 2016). We know that there is a considerable industry in jet manufacture in York (see Allason Jones 1996), but there are also considerable collections of jet and jet-like materials in Colchester and London.

distribution map

FIGURE 2: Distribution of jet object, from Roman Britain, depicting either bears, big cats, or Medusas. Note the strong correlation between these objects and places. 

Upon visiting the museums which have the bears, I was struck by the prevalence on the objects of something I’d overlooked on the figures and drawings of the published references and something which is not addressed in the literature: the presence of incised decoration on the back of the bear. By ‘back’ I mean anatomically – visually this is the upper surface of the object.

img_20180419_133939.jpgFIGURE 3: Jet bear from Bootham, York. Incised decoration on shoulder hump and rump. In the Weston Park Museum, Sheffield.

Two of those from Colchester exhibit this feature: The bear from Abbey field burial CF166 and the bear from the Joselin Collection 87/14 (Crummy 2010, figs. 4 and 5; both currently on display in the Colchester Castle Museum). In the former, the bear generally exhibits more incised markings than the other examples, but at the neck and behind the head are a series of transverse grooves some of which intersect diagonally. The latter is much clearer and has five intersecting incisions.The Bootham bear, now in the Sheffield Museum, is the more prominent of all and depicts four intersecting incisions at both the back/shoulder and the rump.

On the presumption that the bears depicted were modelled on the Eurasian Brown Bear (Ursa arctos arctos), we can see that they are generally good depictions of these creatures and, as far as this particular thought goes, they do depict the strong muscle hump behind the head which is incised on the three examples above. Perhaps the obvious interpretation of these intersecting incised lines is that they depict fur tufts? If I had to wholly oppose this argument it may go down the lines of ‘then why only this space? why not everywhere?’. The bear from CF166, crouching down is generally more incised that the others but most prominently so on the apex of the back. Others are completely smooth and all appear to have been at least partially polished. I don’t think it’s useful to entirely discount this option, as it is one possible interpretation, but an alternative certainly sprung to my mind.

Thus I think it is worth at least suggesting that the depiction of these eight or ten pointed line groups bear strong resemblances to the depiction of stars on other Roman small finds.

YORYM_1972_22_6168 (2)

FIGURE 4: Intaglio depicting stars and crescent moon from the Church Street sewer assemblage, York. In the Yorkshire Museum (YORYM : 1972.22.6168) [CC BY SA 4.0].

Indeed stars are particularly prevalent on specific class of gemstone used for magical purposes and are a a prominent feature of the supernatural iconography on these objects. To see the variety of this feature on magical gems follow this link to the catalogue of the Campbell Bonner Magical Gems Database results for ‘Stars’. They are used as decorative elements on other items of jewellery, such as this metal finger ring (PAS: FAKL-030593). Most prominently they feature in several coin issues, particularly from the 4th Century (e.g. PAS: DOR-B97E44)

The implications of a bear with a star on its back is particularly relevant if we consider the ancient myth in which one of Artemis’ handmaidens, Callisto is transformed into a bear (by Artemis or Hera) after a fling with Zeus which bore her a son – Arcas. Later in the myth the adult Arcas nearly shoots his mother (still a bear) whilst out hunting but both are then changed into stars by Zeus to prevent a family tragedy (Stapleton 1978, 50). Callisto is then the constellation Ursa major and Arcas Ursa minor.

A consideration of this myth opens the speculative door both for thematic links between stars and bears explicitly, but also for parent-child relationships and particularly for the protection of parents and children. Crummy considered the bears as one object type in a broad toolkit used for the protection of infants in the graves in later Roman Britain – they were chthonic objects, placed in the graves of deceased children. Clearly these situations were highly emotively charged, but Crummy’s evidence points to the careful consideration of the most suitable object or (more likely) objects for these situations. Thus the bears may represent part of an explicit link to the loss of a parent-child relationship.

The Greek Magical Papyri (Betz 1992) do not disappoint in continuing the links between magic, ritual practices, and bears: both bears and stars, and the constellation Ursa major were praised and invoked in numerous magical rituals:

  • PGM 1275-1322. “Bear charm which accomplishes everything: Forumula: I call upon you, the greatest power in heaven appointed by the lord gos to turn with a strong hand the holy pole, NIKAROPLEX.”…[continues]
  • PGM 1331-89. “Powerful spell of the bear which accomplishes everything: Take the fat of a black ass, the fat of a dappled she-goat, the fat of a black bull, and Ethipoian cumin, mix all together / and make an offering to the Bear, having as a phylactercy hairs from the same animals which you have plaited into a cord and are wearing in a diadem around you head”…[continues]
  • PGM 686-702 “Bear charm: Bear, Bear, you who rule the heaven, the stars, and the whole world; you who make the axis turn and control the whole cosmic system by force and compulsion;”… [continues]
  • PGM XII. 190-92 “Request for a dream oracle spoken to the Bear: Take olive oil [from] a clean… onto the left hand and say the [names. Then] smear yourself and go to sleep having your head towards thc east. IESOUS ANOUI…”
  • PGM 1-36 “[Rite concerning the Bear: Prepare] an earthen [censer, and] during the 6th hour of the night [offer to the Bear] moss of a savin [in it. Do it before] three days / [after you called on] the Bear nine times from a high roof.”…[continues].

The PGM are clearly referencing the constellation directly, but we may assume at least a passing awareness of the animal to given relevance to the name of the groups of stars in the night sky. These ritualistic elements of the PGM may have direct relevance to the use of Jet in Roman Britain. I have previously argued that certain Jet objects are explicitly and deliberately rubbed in order to produce an electrostatic effect and that this action is evidence of discreet magical practice (see Parker 2016). Several of the bears have certainly been rubbed extensively, perhaps as part of the same effect. It should be pointed out that extensive use-wear of a Jet object in this manner would take an extensive amount of time – rather more than the short lives of the infant burials in which they turn up. Perhaps this is indicative of an object buried with children but used by adults over many more years? Or something curated or used as an heirloom?

Whilst there are only these few scant examples of incised lines upon the backs of bears (3 of the 9 known) is certainly seems plausible that these forms of decorations owe as much to a knowledge and understanding of myth, ritual, and/or magic as they did to the anatomically correct depiction of a bear.



Allason-Jones, L. 1996. Roman Jet in the Yorkshire Museum. York, Yorkshire Museum.

Betz, H. D. 1992. The Greek Magical Papyri in Translation (2nd Edition). Chicago and London, Chicago University Press.

Crummy, N. 2010. ‘Bears and Coins: The Iconography of Protection in late Roman Infant Burial’, Britannia 41. 37-93.

Eckardt, H. 2014. Objects and Identities: Roman Britain and the North-western Provinces. Oxford, Oxford University Press.

Parker, A. 2016. ‘Staring at Death: the Jet Gorgoneia of Roman Britain’, in Hoss, S. and Whitmore, A. (eds) Small Finds and Ancient Social Practices in the North Western Provinces of the Roman Empire. Oxford, Oxbow.

Stapleton, M. 1978. A Dictionary of Greek and Roman Mythology. London, Book Club Associates.

Lucky users

Utere felix – “Good luck to the user”, alternatively translated as “Use, happily” is a simple phrase included on numerous forms of portable object in Roman Britain. It is innocuous, but standardised in its form throughout the material culture on which it appears. It may be abbreviated to UF, but it always stands for the same, single invocation. Variations on this in Greek may be found elsewhere in the Roman Empire.


Base silver finger ring inscribed with Utere Felix. Gussage St. Michael, Dorset.
PAS: DOR-8F5E8E. Image (C)Portable Antiquities Scheme [CC By attribution].

The phrase is close to the modern, Anglophone idea of wishing ‘good luck’ – in modern Britain this is a social nicety, a spoken method of conveying the hope for a positive outcome to oneself or another but one which, if deconstructed, still intends to deliver some sort of beneficial outcome to the recipient. Culturally, in modern Britain, there is a material iconography of luck – horseshoes, four-leafed clovers, and leprechauns feature strongly in this shared understanding of ‘luckiness’. These images are born out of folkloric customs in the preceding centuries and have probably taken on new meaning in this time, but are prevalent today nonetheless. It is probably true also to suggest that there was a material iconography of luck in the Roman world too, though it is much more difficult to access.

It is useful to discuss what ‘luck’ actually is. ‘Luck’ is a concept closely allied with chance, probability, and accident. Darke and Freedman (1997, 488) suggested that there are two main viewpoints on luck (in modern populations): that it is a random and uncontrollable force from which future events cannot be predicted and from which once holds a rational belief about the causal properties of luck, or that it is something which can be embodied in people or objects and remain somewhat stable over time. In attempting to construct a measurable scale of belief in luck they were required to define what they meant by ‘belief in luck’: “Belief in good luck was defined as the view that luck is a somewhat stable characteristic that consistently favours some people but not others and is especially likely to favour oneself (ibid 490).

Van den Hoek, Feissel, and Herrmann (2015) suggest that luck is closely linked to magic in the ancient world because an immaterial power is invoked: “The immaterial force can seem to enter the object [via the text], and thereby becomes a vehicle that acts upon the user. The object becomes personified as an ally who pronounces more-or-less hidden messages of encouragement”. The same sort of functionality may also be true of the icons we describe as ‘apotropaic’ or as serving an amuletic function. The difference is that with messages of good luck, there is no clearly direct reason for needing or using the object – they offer a very generalised sort of benefit to the user.

With Utere felix it is ambiguous whether it is the object itself, the creator of the text upon it, the physical presence of the text itself, or some other exterior force which benevolently bestows good fortune on its user. Interestingly the phrase does not denote ownership – it was not just only the owner that necessarily benefits, but any given user of the object. In this sense the text has an explicitly generalised efficacy. Perhaps the only barrier to using such an object was the ability to read or, at least, to understand the lettering in a meaningful way.

1STY0009 (1)

Copper alloy, enamelled pan with enamelled Utere Felix on the handle. Eastrington, East Riding of Yorkshire. PAS: YORYM-20B68C
Image (C)Portable Antiquities Scheme [CC By attribution].


To date I have catalogued thirty-three such Utere felix objects from Roman Britain: Finger rings (8, including one intaglio), brooches (6), tanks (5), handles (2), spoons (2), strap fitting (1), ceramic vessel (1), bone plaque (1), wooden stake (1). It is evident that at least twenty of these objects are designed for personal adornment. The belt and strap fittings, and the brooches are also clearly designed to fulfil a weight bearing function; they are devices used to hold or affix. It is perhaps an issue of translation and modern semantics whether one can ‘use’ a finger ring simply by wearing it. If its use was amuletic by virtue of the UF text included on them this may be a somewhat circular argument.

It is certainly interesting that many of the object types listed here require physical manipulation to fulfil their utilitarian purpose: brooches must be clasped, patera and spoon handles held, belts clasped etc. and by doing so there was a very physical connection with the text. Not only was it worn, overtly, in these places but it was interacted with by the user to get it into that position in the first place. Perhaps this link between the text and the real world ‘user’ was important?

A differing logic may apply to the lead tanks, the so-called ‘baptismal fonts’ of the fourth century, as the ‘user’ in this sense was someone immersing themselves within its liquid contents; they are also strongly connected with Christian iconography on the exterior of the vessels. The UF phrase predates the rise of the Christian faith in Britain by at least two centuries and, if the semantic argument present above is valid it was a functionally Pagan literary device, so its amalgamation for this purpose is of interest.

The varied nature of the material objects that held the phrase is a point of interest in itself as it shows that the phrase was not necessarily dependent upon specific material conditions to be installed onto something. UF may be a sort of magical panacea in the ancient world; an off-the-shelf, one size fits all sort of apotropaic device capable of being moulded to the needs or desires of the individual.

Perhaps the relatively small numbers of UF devices is somewhat mitigate by the much larger numbers of depictions of Fortuna and Bonus Eventus in Roman Britain – the faces of luck. Of the 2,012 intaglios researched by Ian Marshman in his recently published PhD he found that Bonus Eventus was the most commonly depicted motif in Roman Britain (122 examples, 6%) and Fortuna also featured prominently (75 examples, 3.7%) (Marshman 2015, 140-141, nos. 17.00-17.122, 32.00-32.75).


Darke, P. R. and Freedman, J. L. 1997. ‘The Belief in Good Luck Scale’, Journal of Research in Personality 31. 486-511.

Marshman, I. 2015. ‘Making Your Mark in Britannia: An investigation into the use of signet Rings and Intaglios in Roman Britain’. PhD thesis, University of Leicester. [Available]

Van den Hoek, A., Feissel, D. and Hermann Jr., J. J. 2015. ‘More Lucky Wearers: The Magic of Portable Inscriptions’, in Boschung, D. and Bremmer, J. N. (eds) 2015. The Materiality of Magic (Morphomata 20). Paderborn, Wilhelm Fink. 309-356.


Material Approaches to Roman Magic: A brief reflection.

On this grey spring day, a new book is blinking in the half-light and taking its first tentative steps into the world of academia. This is day for celebration, as an edited volume organised by myself and my co-conspirator Stuart McKie (OU PhD graduate, now lecturer at Manchester University) is released.

Material Approaches to Roman Magic: Occult Objects and Supernatural Substances is an edited volume consisting of ten standalone papers by a mixture of established academics, early career researchers, PhD students, and museum professionals, and weighs in at just over 100,000 words. I thought that now seems to be an ideal time to take a moment to reflect on the process, the finished product, and the experience. There is an official release party in Edinburgh tonight, at the conference, which neither of us are able to attend. Thankfully there will be contributors present to bask in the glory of their newly published work. Writing this blog is how I will be celebrating.

I’d worked as an editor before, but as solo editor, chief people wrangler, reviewer, and type-setter for the British Archaeological Report Ad Vallum: Papers on the Roman Army and Frontiers in Celebration of Dr Brian Dobson (2017). My main memory of that process was that it was a long, lonely one. That, and type-setting 16 papers into Microsoft Word complete with page numbers, headers etc. is a 21st century level of Hell. Material Approaches, affectionately known as the ‘TRAGIC’ project behind the scenes (TRAC + Magic = ….), was a much more positive experience. There are two main reasons for this: Oxbow type-set everything; I had a co-editor to work with.

The inception for this volume was in the 2015 TRAC panel “Charmed, I’m sure: Roman Magic – Old theory, new approaches”. The six original papers included in that session were:

  • Stuart McKie: Distraught, drained, devoured, or damned? The importance of individual creativity in Roman cursing.
  • Glynn Davis: Excessively charged amulets? Contextualising Roman London’s ‘Magical’ Jet and Amber
  • Alissa Whitmore: Fascinating Fascina – Revisiting Roman Phallic Pendants
  • Alessandro Quercia & Melania Cazzuolo – Fear of the dead?Deviant’ burials in the Roman Northern Italy.
  • Idit Sagiv: Victory of good over evil? Apotropaic animal images on Roman engraved gems
  • Adam Parker: Making Magic Work

From this original six, Stuart’s paper on cursing and Allessandro and Melania paper on Deviant burials both made it into the 2016 publication of the 2015 conference proceedings where they sit next to each other. I don’t remember how it happened, but Stuart was quickly on board for co-editor duties, providing a different contribution inspired by his then almost-finished PhD on the topic. Idit’s paper was written up directly for this volume I’m glad to say, and Glynn took an element of his paper and broadened it out – his complete catalogue of the amber from London is forthcoming in the regional journal. Alissa’s subject is also retained in the final volume, though the title and some of the content was used for a paper she published in a 2017 edited volume on dress in antiquity. So far, so good then for producing academic impact from a conference session

The elephant in the room is my own contribution. Originally I had intended to try and bolt some theory together and promote a newer, contextualised approach to the material culture. Whether it worked on the day is for others to judge, but on paper it just looked like I’d bitten off more than I could chew. It took me a few months of occasionally messing around with it to realise that it simply would not work as a standalone paper and certainly not for the volume we were organising, with its material-led approaches. It was a genuine relief to park a paper that I’d found hard to write and move onto what (in my opinion) is what I’m good at – writing about objects. Thankfully, some of what I’d produced was re-worked by Dr. McKie into the introduction, for which he also draws on his own PhD work. The content I’d tried to write makes a lot more sense at the beginning than where I’d originally envisaged it as a paradigm-shifting, world-burning paper as it could never have been either of those things. The choice to write, instead about tintinnabula was drawn from my recent blog post and the experiences of writing some underpinning work on the materia magica of Roman Britain for my PhD research. They are fascinating things and, apparently, wildly understudied as a group of objects.

I’m still not 100% satisfied with my own paper in this volume, probably because it doesn’t really answer any of the questions I’d raised (but nor did it intend to!). It is probably a whole separate research project to fully investigate this topic. There are, of course, many elements of this book which I am very, very proud of. Looking at the list of contributors, the majority of those involved are not the usual big-hitters of Roman material culture/religion/magic studies, though Dasen and Wilburn can certainly be classed as such. Four of the contributors were PhD students when the process started and when they wrote their papers – I hope this volume will be helpful to the careers of these folks as they join the other ECRs from the contents page; having a PhD isn’t a pre-requisite to academic publication though it may often look that way from the outside of the subject or the outset of a research project. The authors are from different backgrounds professionally as well as personally: at least five nationalities are represented here, with a 6:4 male-female gender split. Four of us work also work outside of academia, in the museums & heritage sector.

I’m very proud of the fact that this project arose when I was an independent researcher, just someone interested in the subject. Ok I was trying to get a PhD place at the time (and now have one), but the volume was born out of an interest in the subject, pure and simple, and fuelled exponentially through the inclusion of other interested and interesting individuals. I have a hazy memory of the evening now, but remember sitting in a Leicester pub with various members of the panel drinking for hours and hours during the TRAC party the evening after the session. At that point we were actively reclaiming the “loony fringe” title that Ralph Merrifield has coined for us in the 1980s – I distinctly remember talking about that very phrase. I’m glad that it now also features in the introduction to the published volume.

Of course there were stressful moments and a few late nights of proof reading on a work-night but the darkest cloud which hung over this project, for me, was having to take the difficult decision to remove a contributor from the process. The deadlines were somewhat idealised, but Stuart and I were very accommodating to extensions as long as they did not impact the long-term results; we ended up with a staggered system of papers in, read, edited, changed, returned etc and a simple colour-coded excel spreadsheet to keep track of it all. We reached a point in the process that, despite the constant promises of a contributor to submit their paper, it was evident they were too unreliable to be part of the process and would have been many months behind the other contributors. The email I wrote explaining our situation and to, effectively, boot-out someone who may have jeopardised this project was a very difficult thing to do. I wrote it twice, edited it carefully for the correct tone and then sent a draft to the co-editor and the series editor. The author in question wasn’t happy. To paraphrase them, they returned an email saying that no work on Roman magic would be complete without a contribution on their specific subject. I did not reply, much as I’d have enjoyed angrily hammering the keys, to tell them that I agreed and that was, in fact, the point. Handily, at the same time one of the other contributors (TD) and I had organised a panel on Medicine/Magic for the 2017 TRAC conference. Having seen Nicky’s paper in this panel, we agreed to approach him for a last-minute inclusion into the project. Thankfully, he agreed and worked like an absolute Trojan to get a paper to us in a short window and then to respond to initial comments from the editors within an even shorter window to keep the book on time. The resulting volume is all the stronger for this paper.

This forced removal of a contributor came shortly after one other author withdrew from the process after talking to a PhD supervisor, so there was a point in the process when we were down to 8 papers in the book. The epilogue/round-up by Dasen was a valuable contribution that also came later in the process, partly as a result of these changes early in 2017, but also from the realisation that a collative, ‘finishing’ paper was needed for this volume to make sense as a whole. I’m glad to have had the opportunity to work with the contributors that we did, though I still have not personally met three of them in the real world. It is my fervent hope to work with many of them again on other projects.

There are a lot of transferable skills inherent in editing a book like this. Time management and team-working are the big take-homes, but personal discipline in reading and re-reading when it matters are important, as was maintaining a flexible approach to getting the work done. As no-one is paying any of us to do this, it is a feather in everyone’s caps.

I hope that the journal reviewers are kind to us, but whatever its legacy I’m pleased to have been able to help create a platform for this volume and to push a research agenda I am very interested in.