Draco tauri pater . A new gold lamella from a private collection in Hungary

In 2019 a tiny gold tablet, once folded several times, turned up in a private collection in Hungary. Due to its physical appearance and the layout of the text, the tablet originally had been identified as a ‘ foil with a Greek magical inscription ’ in an auction catalogue. At ﬁ rst glance, however, it becomes obvious that the text was written with Latin letters. Although parts of the text recall Greek and Latin Christian apologists ’ works, the pagan apotropaic language of the amulet is evident throughout. The text itself proposes a major impact of Greek. This paper offers a preliminary report on establishing the text and gives approaches for interpreting the gold lamella.


INTRODUCTION
In 2019, Hungarian art collector Viktor Kiss acquired a small, thin gold plate, which, according to the official documentation, had previously been put up for auctions in the UK and Germany in 2014 and 2017, respectively.The exact location of the discovery, as well as the archaeological context, are unknown to us.In the auction catalogue, the lamella is dated to the 2nd-3rd c.AD, while the text on its surface is identified as Greek.All the external features might lead to a catalogue description as apotropaic pendulum: "Roman gold foil with Greek curse/magical inscription (…) such foils have been stored in small tubes."(Figs. 1 and 2) At first glance, indeed, it becomes immediately apparent that the whole text is written in Latin letters, and some expressions are clearly constructed in accordance with the rules of Latin grammar, but in some other elements, undeniable Greek influence can be recognized.

EXTERNAL FEATURES
The tablet is rectangular: 5.6 cm in height and 3.6 cm in width.Once it was folded roughly 12 times.There are no damages on the surface.Thus, the tablet is fully legible with 12 and a half Fig. 1.Drawing and photo of the tablet (by A. Barta) Fig. 2. The letters of the tablet lines which cover almost the whole surface.Since it is a very fine and thin lamella, the text completely shows through the verso.
Two scientific examinations were conducted in order to describe the tablet as exactly as possible.Firstly, the sticky spots on the surface were analysed by FTIR spectroscopy, 1 by which a certain type of resin, most possibly myrrh was identified.Myrrh was a common part of everyday life in antiquity, had a spiritual significance in the religious sphere and was applied as medicine. 2n addition, sticky residuals may refer to modern restoration procedure, since restorers use similar materials when working with precious metal objects. 3uring the second scientific examination two measurements were carried out with the intention of assessing the main materials. 4It has been found that the tablet was made of gold alloy, namely 85% of gold, 10% of silver, while most of the remainder is copper.The second measurement was carried out from the dark spot at the upper part of the tablet.It showed a slightly higher ratio of silver, therefore the dark spot can be considered as a result of oxidation.In the absence of written sources, it is not yet possible to determine whether it is a natural alloy or it was produced intentionally (i.e., following instructions) as an alloy.

PALAEOGRAPHY
Although the faint hope of assessing the production date, location, or the archaeological context through scientific analyses failed, still, palaeography may provide some results since the letterforms seem very characteristic.As opposed to what the catalogue claims, they are clearly not Greek.Generally, gold lamellae from antiquity are expected to bear rather Greek text than Latin, but for certain reasons the latter could be adopted.At this moment, the reason is not obvious, but the lamella seems to have been written in Latin script.
The text is a combination of various scripts: cursive and capitalis of epigraphic evidences and extraordinarily, bookhand uncial and semi-uncial: the left side of the A forms a sharp angle and it resembles a Greek alpha; B was written with only a lower bow to the right, which is a characteristic of new cursive and minuscule scripts;5 D has a straight shaft and a sharp bow to the left, tending below the baseline; E was applied in three versions, a capitalis, a C-shaped cursive with one horizontal articulus, and a rounded uncial which resembles a Greek epsilon; F's vertical articulus descends slightly to the left, while the two horizontal shafts tend firmly upwards; the second character in line 11 resembles a carelessly written, curly G with an elongated tail, or it can even be considered a kind of negligent transliteration of a Greek zeta; if the interpretation is correct, the letter H appears twice, and both letters suggest an unskilled hand: at the end of line 7 it was written with uncertain strokes (and it might as well recall a letter R with a very small upper bow), and in line 9, it simply falls into pieces; the M is a roundish letter, resembling the double arch of the bookhand uncial M. In the early version of this letter, two equal arches were added to the initial vertical shaft. 6Here, it was written with three strokes, but the initial shaft is a curve bending inwards.In line 8 probably a mistakenly written letter was corrected into an M, that is why it looks different.N, P and R have roughly the same form as in capital script, and the latter was corrected from an O in line 4; S takes a cursive form with a short oblique stroke on the upper part; the letter occupies both the upper and lower space; the vertical line of the T is straight; V represents a minuscule form, with a shaft that curves to the left, and with a vertical shaft that in some cases runs through the middle of the letter.
As it is usually the case with this type of textual evidence (probable inscription that was produced to do harm or to protect, aiming as precisely as possible, giving no room for misunderstanding), ligature was employed only once, at the end of line 2: in SE the E was marked by only the middle vertical stroke and it was attached to the S in this manner.
The tablet shows an inconsistent overall picture.At first glance, the last three lines do not give any meaningful text, with the words failing to come together.Letters are falling apart already in the previous line 9, but that line is still readable.This inconsistency could be attributed to negligence or lack of skill.The text may have been copied from a draft or formulary book which was not entirely comprehended by the copyist.
On palaeographic grounds, it is almost certain that the original text was created after the 3rd century AD, since uncial letters evolved at that time at earliest.However, some firm, definite letterforms suggest later dating.Kε is a common spelling mistake reflecting the monophthongisation of αι, already from the classical period. 8It is well attested in epigraphy, like in Greek defixiones 9 or presumably also in a Latin defixio, as well, produced among Greek-named slaves. 10In addition, καί was transcribed as CE in Greek codices written in Latin script. 11It is not only the Greek κ that was rendered as Latin C: the evidence of the transliteration Greek χ as C can be clearly seen in the case of CALCV. 12The ending might be a genitivus materiae preceding SIMBOLVS in the following line.In standard Greek it could be either Xαλκoῦ καὶ σιδήρoυ σύμβoλoς (rather σύμβoλoν), or Xαλκoῦ<ς> καὶ σιδήρoυ<ς> σύμβoλoς, where the two adjectives χάλκεoς and σιδήρεoς are meant to be interpreted with a contracted εo, and with a not indicated final -ς.

DIPLOMATIC TRANSCRIPTION
On the other hand, there is some evidence to suggest that Latin interpretation should not be rejected entirely, yet it is much less probable.Looking for a Latin interpretation for CALCV, the 7 The two metals were mentioned commonly together, e.g. on the architecture of Tartarus Il.VIII 15 ἔνθα σιδήρειαί τε πύλαι καὶ χάλκεoς oὐδός 'iron gates and bronze threshold'.Psalterium duplex cum canticis … by Giuseppe Bianchini from 1740 is a good example.This print version of a 5thcentury manuscript with uncial letters from Verona provides a two-page long facsimile of Psalm 142 both in Greek and Latin, each of which is written in Latin script: e.g., 2 ce me iseltes is crisin meta tu dulu suet ne intres in iudicio cum servo tuo (https://www.digitale-sammlungen.de/de/view/bsb10326493?page52) "Do not bring your servant into judgement".most plausible solution could be c al ıg o, either with a G that has become voiceless or marked simply by a C, and an o/u fusion 13 in the very same word,although it is quite unusual to see a syncope in a stressed syllable.But the Appendix Probi points out (124.Caligo non calligo with an L gemination) that in the late imperial period the pronunciation of the word changed: the stress was shifted from the second to the first syllable, so the syncope in CALCV could have taken place. 14A possible explanation for SIDERV is that it is a genitive plural of sidus 'a heavenly body, star, planet' where the final -m was not indicated again.While the two words caligo and sidus occur together in several authors, 15 the interpretation of CE as a Latin word is more problematic, for there is only one dubious record of such a spelling mistake in the most probable reading, in the relative pronoun. 16In standard Latin it could be: Caligo quae siderum (symbolus).
2 SIMBOLVS: A word of Greek origin.According to the text on the tablet, here it means 'objects used in the ritual (such as signs, tokens)' 17 and 'passwords among initiates'. 18In Christian context it means 'creed' (e.g., Symbolum Nicaeno-Constantinopolitanum, Σύμβoλoν τῆς Nικαίας Nicene Creed).As regards the -us ending, this spelling was likely meant to be a Latinised one, 19 although the preceding line can be interpreted much better in Greek.The transcription of the Greek Y is in line with the general trend in imperial inscriptions for the 1st to 5th centuries. 20-3 ROSEV FOEBOS: The present spelling of Φoῖβoς/Phoebus is attested several times in epigraphic evidence. 21The -os ending is a common feature in words of Greek origin and in areas of Greek substratum, adstratum.However, inconsistently, the previous word simbolus was provided with a Latin ending.Classical authors, both Greek and Roman, use the epithet Φoῖβoς/ Manilius, Astronomicon IV 528-532: At, niger obscura Cancer cum nube feretur,/ qua velut exustus Phoebeis ignibus ignis/ deficit et multa fuscat caligine sidus,/ lumina deficient partus geminamque creatis/ mortem fata dabunt.But when that part of the Crab rises which is dimmed by a sombre cloud, where his own fire fails, as though burnt out by the Sun's, and darkens the signs with impenetrable fog, the sight of those born then will fail, and fate will condemn them to death twice over (transl.by G. P. Goold).Ovidius, Met.I 69-71.Vix ita limitibus dissaepserat omnia certis,/ cum, quae pressa diu fuerant caligine caeca,/ sidera coeperunt toto effervescere caelo.He had barely separated out everything within fixed limits when the constellations that had been hidden for a long time in dark fog began to blaze out throughout the whole sky (transl.by A. S. Kline).Phoebus resp.when they refer to Apollo, in his capacity of God of the Sun.On the other hand, it should be mentioned that, in the official iconography with the exception of the classical Latin authors who go back to Hellenistic patterns, the identification of Apollo with the sun god begins only during the reign of Augustus. 22Phoebus' epithet roseus 23 originates also in classical authors.One question that needs to be asked, however, is what the reason for this word choice was, why Phoebus (or Apollo) was invoked with an epithet referring to dawn or sunset. 24As for the spelling, a frequent vulgar Latin phenomenon, the loss of the final -s25 is distinctly visible in ROSEV, which form here might have been supported by the words ending in -u in line 1.Moreover, the only ligature of the text occurs in this word, affecting the letters S and E.
When we compare the expression at issue with the above loci, the most plausible interpretation, instead of the more obvious Latin version ( p tauros draconesque, draco tauri pater) runs as follows: taurus draconis kai (5et) draco tauri pater.Greek and Latin linguistic elements are mixed: taurus is a Greek loanword in Latin, but the present form TAVROS must represent a Grecised nominative singular, in contrast to SIMBOLVS, but similarly to FOEBOS.DRACONES pro draconis is a genitive singular of another loanword draco already built in the classical Latin vocabulary.Writing ES instead of the genitive -is proves the most common vocalic fusion, namely that E is written where ȋ is expected. 26The καί of the Greek originals by Clement of Alexandria and Firmicus Maternus was written here again as CE like in line 1.The second part of the expression is of Latin morphology.

24
The legend of Augustus' conception (Suet.Aug 94) which applies similar vocabulary, must be approached here with some caution, since it was a story approved (or even supported) by the imperial power in contrast to the possible context of our tablet, the secret mystery cults or even magic.However, it is of special interest in the view of the following expression on snakes: before the conception of Augustus (who increased the god's importance in the Roman religion), a snake crawled into the lectica of his mother (ob hoc Apollinis filium existimatum), his father then had a dream that the sun rose from Atia's womb (Somniavit et pater Octavius utero Atiae iubar solis exortum.)Augustus' cult of Apollo is generally considered a counterpart to Antonius' cult of Dionyus, to whom the following lines of the tablet refer (cf.n. 22).
Since the works by Clement, Firmicus Maternus, and Arnobius were read throughout the following centuries of Christianity, their treatises on contemptible mystery cults have gone down to posterity.Quite a few further authors make references to them or simply cite paragraphs.One of these works is the Lexicon Universale published in Geneva, in 1677 27 which mentions the Greek saying, and provides its Latin translation, both in a word order which is in line with the one on this tablet: "Apud utrumque, ταῦρoς δράκoντoς καὶ δράκων ταύρoυ πατὴρ Taurus draconis et draco tauri pater." 287 IAO SABAOATH| (or SABAOATR|): It is not clear yet which spelling of the great god of Jews was meant to be employed here, and where the words can be separated from each other.The most common form is Iao Sabaoth, but further instances also attest the Sabao spelling variation. 29n both cases, after the letter O, a superfluous A is assumed.At the end of the line, the letter H may be explained as R, too, since a small bow is discernible at the top of the vertical articulus.
8 |AXEPI: Without postulating any spelling mistakes, these letters are difficult to interpret on their own.Due to the words daemonium and phantasmata, a possible explanation might be a corrupt form of Acheron, if the numerous inconsistencies do not raise serious objections: 30 in this way, the letters X and P should probably be explained as Greek chi and rho, while the final -i must be a Latin substandard genitive. 31When summoning deities in magical papyri, requests are made with the words τάχoς ἐπὶ (γαῖαν). 32If we accept this as a possible reading, unexpectedly the loss of the final syllable -oς is clearly presumable; moreover, the last two letters of the previous lines representing a [t] are needed, as well. 33Further corrupt solutions, spelling mistakes displaying phonetic changes, or simply the negligence of the scribe, could be revealed. 34ut in lack of direct parallels, no convincing interpretation can be offered.For Acheron, the Latin genitive -i is never attested.In Latin inscriptions both standard Greek and Latin genitives were written: ego non dubitem tacitis Acherontos in umbris (CIL VIII 00212, p. 925, 2353), in tenebris Acherontis labitur undis (CIL X 08131).
8-10 DEMONIVM E PHANTASMATA: Greek has several all-purpose words for apparition or ghost, such as δαίμων, δαιμόνιoν, φάσμα, φάντασμα and so on, generally without any specific denotations. 35A magical papyrus preserved the instructions how to make a protective amulet against these kinds of ghosts. 36As regards spelling here, while the Greek φ is rendered by a Latin F in Φoῖβoς, the Greek plural of φάντασμα is transliterated with PH in Latin; however, the word itself is hardly recognizable.In DEMONIVM, E stands for the Greek αι or the Latin ae, representing the monophthongisation of the diphthong, whereas the ending -ium clearly proves a Latinised form such as SIMBOLVS above.The E between the two words must be the remainders of a conjunction stemming from either the Greek καί > CE > <C>E, or the Latin et > E<t>.
11-12 IGITESCEOMA|CARIIIN: The letters in the last three lines have failed to offer a convincing meaningful reading so far.The second character at the beginning of line 11, possibly a G, may be a letter of ἡγέτης/ ἡγητής, an attribute of gods leading various groups. 37However, caution must be applied, as Greek Z could be written with similar strokes, and presuming it in this place, further words must be considered. 38There is a corrected letter between E and M: maybe O is the final one.At the end of the line, the A was added later, in a smaller form (probably should be read as καὶ ὄμμα?).The character at the beginning of line 12 looks more like a C than T written by an unskilled hand because of the lower curve bending to the right.The next character is clearly an A, but the following letters have fallen apart which raise difficulties in the interpretation.
12-13 CASTINI: Made up of well readable letters, the last word of the text evokes ritual purity in Latin.Still, the above lines based on Greek originals may give rise to doubts whether the obvious Latin name of the owner was meant here; or rather it needs to be interpreted as a Latin version of καθαρμός, a term in the context of mystery cults and magic.

INTERPRETATION
The present gold lamella is of much interest since it represents characteristics of various kinds (let us say, genres) of textual evidence from antiquity.Gold tablets were usually produced in order to serve as protective amulets for those alive or Totenpasses for the dead.Although outer and inner featuressuch as the use of magical kharakteres and vocabularymay help classify these texts, in the case of the lamella at issue, these features have failed to offer clear clues so far.
As for external features, the letterform resembles bookhand uncial in contrast to the cursive, capitalis and minuscule scripts of other tablets.Although metal amulets span a large period of classical antiquity, 39 the letterform of this lamella suggests a later date.
The language of the tablet might also indicate a later date: the text is clearly inscribed in Latin letters, and certain parts can definitely be interpreted in Latin.However, Grecisms suggest that it was originated in a Greek "draft" or formulary book.Most precious metal tablets, even in the Latin part of the Roman Empire, were written in Greek, possibly mixed with nonsense voces magicae. 40Magical papyri (considered as sources for phylakteria) are mainly in Greek, rarely Old Coptic and Demoticpartly due to the provenance.This source can be dated roughly to the same period. 41At the end of antiquity, magical beliefs and religional beliefs other than Christianity had not disappeared entirely, but the practice had changed in many respects.The most important of these for us is the language: Latin became prevalent in the western part of the former Roman Empire.
From the very beginning of the tablet, the reader cannot help noticing that it probably comes from a Greek original, which seems to have been forcibly and inconsistently inscribed in Latin and in Latin letters.The reason for this is not yet clear to us.Certain words and expressions of the tablet can be explained either in Greek (such as CALCVCESIDERV from χαλκoῦ καὶ σιδήρoυ) or in Latin (draco tauri pater).Yet, the peculiar use of CE raises the possibility that the tablet may have been made from an original in which the Greek text was written in Latin letters, and that the author of the tablet, when copying the text onto the tablet, Latinised the words he figured out (or rather, we may say, he adjusted it to Latin), but left the rest in their original Greek language form.The hypothesis works well for SIMBOLVS, DEMONIVM, DRACO TAVRI PATER, and even TAVROSDRACONESCE, where he may have "resolved" the TAVROSDRACONTOS form of the supposed original as plural accusatives in Latin and thought they might have been joined by the -que particle, written as CE.
Thus, the generally accepted reason for code-switching in magic, that is the concept "magical texts are written in more than one language presumably to increase the obscurity and enhance the power of the magic" 42 appears to fail here.This plate seem to have been produced by an otherwise skilled person with incomplete knowledge of language and traditions of the genres that gold tablets represent. 43evertheless, there are still some interesting and relevant problems to be addressed.papyri, as demonstrated above.The instructions of PGM VII 579 51 are clearly in line with this tablet.On the one hand, it prescribes to write either on a leaf of gold, silver, tin, or hieratic papyrus.On the other, it works with the name of the great god (ἔστιν γὰρ δυνάμεως ὄνoμα τoῦ μεγάλoυ θεoῦ; Iao Sabaoth from line 7 of our tablet, or perhaps the god hidden in the saying draco tauri pater and his supplemental counterpart, Phoebus, as well?)In addition, Iao Sabaoth (with this form of his name) is more related to the context of magic.Nonetheless, in various sources he was identified with Dionysos/Zeus Sabazios.52

CONCLUSION
While this study does not offer a comprehensive description and interpretation of this newly discovered gold tablet, it does give approaches.From the research that has been carried out, it is possible to conclude that the tablet must have been produced after the 3rd century AD, perhaps much later, in a Latin language ambience with Greek backgrounds.To determine whether the text was created in a context with roots to mystery cults or protective magic, further investigation is needed.

ACKNOWLEDGEMENT
I thank audiences at Budapest, Gent, and Vitoria for their insightful comments after I delivered oral versions of my work in process; I am also thankful to the reviewers for their suggestions that helped to improve the manuscript.Any errors that remain are my sole responsibility.The present paper was prepared within the framework of the NKFIH (National Research, Development and Innovation Office) project no.K 135359 entitled Computerized Historical Linguistic Database of Latin Inscriptions of the Imperial Age (see: http://lldb.elte.hu/).I wish to express my gratitude to Gábor Lassányi PhD for drawing my attention to this amulet, and to Judit Császár for her help in the revision of the English text.
Open Access.This is an open-access article distributed under the terms of the Creative Commons Attribution 4.0 International License (https://creativecommons.org/licenses/by/4.0/),which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited, a link to the CC License is provided, and changesif anyare indicated.(SID_1)

51
"A phylactery, a bodyguard against daimons, against phantasms, against every sickness and suffering, to be written on a leaf of gold or silver or tin or on hieratic papyrus.When worn it works mightily for it is the name of power of the great god and [his] seal, and it is as follows: (voces magicae) These [are] the names; the figure is like this: let the Snake be biting its tail, the names being written inside [the circle made by] the snake, and the characters thus, as follows:
Despite the Latin letters, a Greek interpretation is much more likely, which may refer to the two metals χαλκός -σίδηρoς, copper/bronze and iron, 7 separated by CE.

39
Magical amulets date from the 2nd c.BC to 6th c.AD (KOTANSKY, R.: Greek Magical Amulets: The Inscribed Gold, Silver, Copper, and Bronze Lamellae Part I Published Texts of Known Provenance.Opladen 1994, XVII-XIX, while the socalled orphic gold leaves date from 400 BC to 260 AD (BERNABÉ, A. -JIMÉNEZ SAN CRISTÓBAL, A. I.: Instructions for the Netherworld: The Orphic Gold Tablets.Boston 2008, 2).