1As the first of a pair of contributions¹ to this volume, I am delighted to publish the first of the two parchment sheets discovered among the papers of Walter Ewing Crum in the archive of the Griffith Institute: a figural amulet or deposit depicting the mythical precedent of Apa Baula’s triumph over The Destroyer (Satan) to either protect a client or bind a target.

Discovery in the archive

2One rainy summer’s afternoon in 2012, Archivist Cat Warsi and I were in the archive of the Griffith Institute leafing through the papers of Walter Ewing Crum.²

3We came to C/Group XXIII.2 – labelled “Original Papyrus + Paper fragments”, the contents of which were housed in a simple brown envelope slightly bigger than A5 in size. The annotations “2 parchment fragments, magical, bought Cairo 1905 by Petrie”, among others, leapt out at us. With intense but measured curiosity, I removed an old newspaper sheet of 24.2 x 10.4 cm (Crum xxiii_2_3b, fig. 1) from the envelope, and saw written upon it “bought Cairo ’05 F. Petrie” – that is, Flinders Petrie, considered one of the founding fathers of modern archaeology. Dating to Thursday 4th May 1905, the contents of this copy of The Standard itself made interesting reading, but what was folded within it was an even greater curiosity: two parchment manuscripts.³ Petrie presumably sent these to Crum because he thought they would be of interest to him, although it does not appear from Crum’s papers that he ever began work on them. One was unfolded, and facing us as we opened the newspaper,⁴ while the other was folded (Crum xxiii_2_1, fig. 2 and 3), obscuring its contents, except for the dark ink traces that could be seen through the hair side of the parchment. Marvelling at the ‘demon chicken’ facing us (Crum xxiii_2_2, the subject of my subsequent contribution), we then prised the second parchment gently apart, and thus we were able to discern the outline of its figures and inscriptions. After Crum xxiii_2_1 had been relaxed and appropriately rehoused, the tableau of figures could be properly appreciated. It was not long before the principal characters of Apa Baula (ⲁⲡⲁ ⲃⲁⲩⲗⲁ), the name above – and presumably labelling – the anthropomorphic figure, and The Destroyer (ⲡⲣⲉⲃⲧⲁⲕⲟ), the name above the zoomorphic figure, were identified. Thus, the search for their story could begin – together with the relevance of their accompanying menagerie of zoomorphic and composite figures, voces magicae (“magical names”), and pronounceable but unintelligible charaktēres (“characters”) inscribed upon the parchment.

Crum xxiii_2_1

4The figures, signs, and text of Crum xxiii_2_1 are inscribed upon the relatively clean flesh side of a light brown/dark cream parchment sheet whose hair side has been left blank. With the exception of the top right-hand corner of the inscribed recto, where the corner has been cut off diagonally, the parchment is rectangular, with sides that were cut relatively straight. With a height of 21.3-22.5 cm and width of 16.2-17.3 cm, the parchment is twice as high and almost half again as wide as Crum xxiii_2_2.⁵ When folded along its horizontal creases it would have produced a package of 5.5-5.7 cm in width and 21.30-22.50 cm in height – an oblong shape that would have accommodated Crum xxiii_2_2 within it.⁶ When folded along the traces of its vertical creases it would have produced a package of 5.5-5.7 cm in width and 5.4-6.1 cm in height. The parchment is remarkably well preserved, but exhibits patterns of damage along its top and bottom edges where a couple of centimetres in width have been eaten away, as at the body of Apa Baula. Along the central horizontal crease, as well as at the intersection of the top and bottom horizontal creases with their respective rightmost and leftmost vertical creases, the parchment has been worn and eaten away, resulting in nucleated holes along these creases. Inscribed with an ink of a relatively dark brown colour,⁷ this hue resembles that of other late magical texts on parchment, in particular those now in the Institut für Papyrologie der Universität Heidelberg.⁸

5The pattern in which the folds proceeded (see table 1) can be discerned from the traces of the ink across these folds, resulting from the ink still being wet when the parchment was folded, which in turn corresponds to the relative weight of those creases as exhibited on the verso. Firstly, the bottom⁹ and top¹⁰ edges were folded inwards until they met each other. Secondly, the parchment was folded in half again from its new bottom edge upwards towards its new top edge, producing the central horizontal crease. As with the horizontal relative to vertical creases of Crum xxiii_2_2, the vertical creases of Crum xxiii_2_1 are very faint relative to the horizontal creases, thus it is less clear how the folding proceed. However, the little damage exhibited on the right and left edges suggests that these edges were then folded inwards until they met each other. It would then proceed that the new right and left edges were folded inwards again, producing the central vertical crease. If so, this would account for the weight of the darkened creases visible on the verso, as well as the most considerable patterns of damage, such as the eaten-away body of Apa Baula, the verso of which would have formed part of one of the outermost faces and creases of the folded package.

6Inconsistent with this, however, is the damage along the top and bottom edges, which cannot be easily reconciled with the ink traces that suggest the parchment was first folded inwards from its bottom and top edges into the centre, leaving those top and bottom edges within the folded package. Although these patterns of damage are focused on the edges of vertical creases, they do not mirror one another, in turn all but ruling out the possibility the parchment was unfolded and folded again along the central horizontal crease.

7If the initial pattern of folding described, consistent with the ink traces exhibited on the recto and relative weight of creases exhibited on the verso, is correct, this would have produced a package that adheres to the millennia-old Egyptian tradition through which no single outer edge is exposed and thereby the ‘contents’ of the package are sealed within – a small oblong package that was then worn or deposited.

8Given there is very little Coptic text on Crum xxiii_2_1, it is difficult to suggest a dialect in which the readable words are written. There are, however, two diagnostic features which can be treated. The orthography of the name Paulos (Παῦλος) varies over spatial and temporal contexts – even within Egypt there are more than 30 attested orthographies¹¹ – but ⲃⲁⲩⲗⲁ is not yet attested in the Trismegistos database. The orthography with ⲃ rather than ⲡ, which reflects /b/ < /p/ as a result of Arabic influence, and the final vowel ⲁ, also suggestive of Arabic influence, suggest a temporal, rather than spatial, context. By comparison, in the name ⲡ-ⲣⲉ=ⲃ-ⲧⲁⲕⲟ, the orthography of the pronoun with ⲃ rather than ϥ is a feature found in “especially Fayyumic and non-literary Sahidic” Coptic texts,¹² but also dialects N and H¹³ – the former of the 10th-11th century from the region of Oxyrhynchus and Ashmunein and the latter from Hermopolitan region of the 9th to as late as the 10th/11th century. ⲧⲁⲕⲟ, however, is the standard Sahidic and Bohairic literary orthography.

9Dating Crum xxiii_2_1 through palaeography is first and foremost problematic due to the ‘informal’ nature of such a text compared with the ‘formal’ nature of those texts which constitute the majority of examples in Coptic palaeographies. In addition, with limited text, and the only diagnostic letter morphologies being those of ⲩ¹⁴ and ϥ,¹⁵ as well as the ⲩ-derived charaktēr¹⁶, the palaeography cannot provide a definitive date. The tableau itself, however, can provide supplementary information for dating based on comparable examples in other more securely dated/datable manuscripts. For example, the morphologies of the facial features of Apa Baula and the ‘owls’ are similar to those common in the figures of the 9th and 10th century Coptic magical texts¹⁷ and distinct from anthropomorphic examples from some earlier examples.¹⁸ However, such morphologies are also found in certain figural traditions from 6th/7th century Coptic-language¹⁹ and 5th/6th century Greek-language²⁰ examples. Taking into account the orthography of the name of Apa Baula (i.e. Paulos), one could easily suppose a date after the conquest of Egypt, and therefore a temporal frame from the 8th to 10th centuries²¹ seems most viable. Most recently, however, the dating of several comparable manuscripts from the so-called Heidelberg Magical Archive has been made possible by the identification of a colophon written in 967 CE by the Deacon Iohannes.²² Given the similarities between the hand of, and the morphologies of some of the figures in, those manuscripts and Crum xxiii_2_1,²³ a contemporary dating in the (second half of the) 10th century seems most favourable.

The tableau of Crum xxiii_2_1

10Descriptions of figural content are inevitably subjective, while following such descriptions can be tiresome for a reader. A close consultation of figures 2 and 3 is therefore encouraged. However, given that my interpretation of the tableau of figurae magicae, voces magicae, and charaktēres has formed the basis of my research into the Crum parchments, it is still prudent to present this as a starting point, following an ‘edition’ of the textual content of this amulet. Some aspects of the description that follows may therefore be insightful to a reader, whereas others will be subjective, described in arbitrary language, and thus potentially unsatisfactory.

Crum XXIII_2_1

24.2 × 10.4 cm
Provenance unknown
Parchment
x CE

Figures

11Below column one are several squares exhibiting protrusions parallel to their framing lines, which at first sight resemble rugs or mats.⁴⁰ Each of these, with the exception of the rightmost example, features a letter-derived charaktēr attached to it – from left-to-right ⲯ̣, ϫ̣, ⲓ (horizontal), ⲁ, ϯ/+. At the bottom left-hand corner of the parchment is a front-facing figure which resembles an ‘owl holding a banjo’, with a ⲑ-derived charaktēr under its left arm and an ⲓ-derived charaktēr on its torso. This figure, like that above it and to its right, has a face that, with the exception of the absence of a mouth, is very similar to that of Apa Baula. While the leftmost of these ‘owls’ has a hand with three digits at the end of straight arms and nothing upon its head, the rightmost has wings and a four-pointed cross as well as one vertical line upon its head, both of which terminate in rings. In addition, the leftmost appears to terminate at its base with a horizontal platform which is vertically striated, while the rightmost exhibits undulating lines extending diagonally from each bottom corner of the torso, a torso which – by comparison – is not inscribed with any charaktēres. Underneath the rightmost ‘owl’ is a rhomboid terminating at each point with an equilateral triangle, this is comparable to the rhomboid underneath the anthropomorphic figure of Apa Baula, terminating at its top and bottom with an equilateral triangle, but at its corners with rings, which is thus six-pointed. To the right of the rightmost ‘owl’ is an L-shaped avian figure in profile, inscribed with one single circular eye, a beak, one anthropomorphic arm – comparable with that of Apa Baula – and two anthropomorphic legs and feet facing in the same direction as the figure itself. To the right of this figure is an eight-pointed star terminating in rings, and under this a charaktēr also featuring eight rings, but in the arrangement of a vertical stem, with two horizontal protrusions, and one hemispherical protrusion at its top.

12The composition of the tableau as a whole is dominated by the anthropomorphic and zoomorphic figures labelled ⲁ̅ⲡ̅ⲁ̅ⲃ̅ⲁ̅ⲩ̅ⲗ̅ⲁ̅ and ⲡ̅ⲣ̅ⲉ̅ⲃ̅ⲧ̅ⲁ̅ⲕ̅ⲟ̅ ⲥ̅ⲁ̅ⲣ̅ⲁ̅ⲧ̅ⲁ̅ⲗ̅ⲏ̅,⁴¹ respectively. Apa Baula is depicted front-facing, with anthropomorphic facial features and seven vertical protrusions from his head, terminating in rings. His torso, although mostly lost, does not possess legs or feet, but is seemingly clothed, and is inscribed with letter-derived charaktēres, while his two arching arms terminate in five-fingered hands. ⲡ̅ⲣ̅ⲉ̅ⲃ̅ⲧ̅ⲁ̅ⲕ̅ⲟ̅ ⲥ̅ⲁ̅ⲣ̅ⲁ̅ⲧ̅ⲁ̅ⲗ̅ⲏ̅ is a complex composite figure with a head and face, which, when the parchment is turned on its leftmost edge, would appear front-facing, but whose body and limbs are in profile. The face differs from that of Apa Baula and of the ‘owls’ in its morphology, featuring a snout/muzzle exhibiting vertical strokes – presumably teeth – which terminates in rings at its corners, and has two pointed ears which stand diagonally upwards from the head, with a seven-pointed charaktēr as a headdress between them. The neck is long in proportion to the head and body, and appears to be gripped or bound by a straight horizontal protrusion from Apa Baula’s arm. The morphology of the body is that of a rounded rectangle joined on the left to Apa Baula’s arm by a feature which looks like a spinal column, as well as a straight diagonal protrusion which then extends from the bottom left of the body and terminates in an equilateral triangle with rings at its vertices. Above both of these connecting lines is inscribed a symbol which may be a cryptogram. The body is inscribed with charaktēres.⁴² On the top of the body is an equilateral triangle, while the rear of the body features a very long tail that curls upwards and is flanked in the first half of its length with 13 straight protrusions which extend parallel from the tail, accompanied by three ⲱ-derived charaktēres. In the second half of its length, nine horizontal lines bisect the tail, terminating in rings, before the tail itself terminates in a three-pronged fork ending in rings. Eight of the nine legs were drawn in pairs, the third pair of which appears to be bound around the ‘heels’, all of which terminate in single rings – presumably hooves – while the ninth terminates in a three-pronged fork itself terminating in rings – identical to that of the tail. Between the second and third pairs of legs features what may be its teat, below which is written ٥١٤, perhaps the Arabic numerals 514,⁴³ and a ⲫ terminating at its base with a ring,⁴⁴ inscribed underneath.

Figural amulets from Egypt

13Inscribed material with ritual applications for the benefit of a living individual, i.e. sources attesting to practices conventionally and conveniently termed “magical”, from Roman, Late Antique, Coptic, and Islamic Egypt, can be categorised primarily based on their form: (1) figural and (2) textual amulets, (3) loose leaves and loose-leaf portfolios,⁴⁵ and (4) bookrolls or codices,⁴⁶ as well as their application: (a) master copies (formularies) or (b) applied/activated texts (produced for a client). Crum’s parchments belong to the category of “figural” amulets and are likely applied/activated. Figural amulets are defined here as those which consist primarily of figural elements, i.e. figurae magicae (“magical figures”). While they may contain textual (i.e. ‘written’) elements – voces magicae⁴⁷ and charaktēres⁴⁸ –, these do not encode phonetic, semantic, and syntactic elements in the same way as written language. By contrast, “textual” amulets⁴⁹ are those which contain textual element(s) which encode a written language. Thus this definition does not preclude the idea that figural amulets could indeed be ‘read’, but it does reflect the fact that they do not encode ‘meaning’ through the normal conventions of written language.⁵⁰ These two formal definitions are not black and white categories, but opposite ends of a spectrum of possibilities: figural amulets without figures⁵¹ may be compared with textual amulets which are predominantly figural.⁵² Thus, by analogy with the model of “textual authority”,⁵³ such figures and their accompanying inscribed elements could be conceptualised as encoding a “figural authority”.

14Crum xxiii_2_1 does not feature the instructions required of the ritual practice accompanying its production, and therefore the content of the amulet does not itself reveal whether it was deemed inherently efficacious or if it had efficacy applied to it through such a practice. As a result, this contribution is not a simple ‘text edition’, but instead a study which engages with questions arising from the ‘writingless’ (non-syntactic) tableau of Crum xxiii_2_1 by treating: (i) figural parallels in the contemporary Coptic magical text tradition, and (ii) the preceding tradition of the Graeco-Egyptian magical papyri – comprising the so-called Demotic and Greek magical papyri (PDM and PGM)⁵⁴ – as well as (iii) the “graphic reservoirs” of the “figural culture” – the art and iconography of the preceding and contemporary milieus of Coptic and Islamic Egypt – which may have provided sources for transmitting novel figures into the magical text tradition.

Prior scholarship on the figurae magicae of magical texts from Egypt

15When beginning to investigate the tableau of Crum xxiii_2_1, I started by considering what parallels could be found and how the phenomenon of so-called figurae magicae had been interpreted thus far. Discussions of selected magical figures and their relationship to the texts which they accompany are often found,⁵⁵ although little has been produced until recently that treats them in their own right.⁵⁶ Gordon sees them as “proleptic images”, “secret, unseen by anyone except the practitioner and his addressee(s)” which suggested “that it was precisely in malign magic, generally considered the most difficult to perform, that specific efforts were felt necessary to increase the force of the text”.⁵⁷ Yet, if it were simply a matter of increasing the efficacy of a text, whether inscribed or recited, both of Crum’s parchments – as figural amulets – must have had either applied or inherent efficacy. In the most recent study of magical images, the authors concluded that both “text” and “image” should be seen “als festes Schema/Formular”.⁵⁸ However, while such revisionism is important, given that the figures themselves have been chronically underappreciated, or even derided in the majority of the scholarship of the last century, a revisionist approach that reacts to the ignorance of figures with the insistence that they are inseparable from the text is itself erroneous,⁵⁹ and therefore counterproductive.⁶⁰

16Hence, more typological research is required,⁶¹ and research which takes into account the fact that figures, known as vignettes, are ubiquitous in the figural and textual culture of Pharaonic and Graeco-Roman Egypt. They are not, as will be summarised below, an invention of the Graeco-Egyptian syncretism of the Roman Period for utilisation in magical texts, but the adaptation of a millennia-old practice of similar forms to different functions.

Figural amulets and magical figures in the Egyptian tradition

17The fact that adding vignettes to ritual texts comes from Egyptian tradition is taken as a given by Egyptologists, but does not appear to have made much headway in the wider discourse on the Graeco-Egyptian magical papyri or Coptic magical text tradition.⁶² In contrast, such an influence has been taken as a given for the influence of Egyptian on Graeco-Roman art and iconography. Hence, there are two possible explanations: (1) the tradition of utilizing figures was transmitted directly to the Graeco-Egyptian magical papyri, and subsequently into the Coptic magical text tradition, from the hieratic and Demotic magical and afterlife texts of the Graeco-Roman Period; (2) the tradition of utilizing figures was emulated⁶³ by copyists of Greek magical texts who themselves had no direct link to the earlier culture, before being subsequently transmitted into Coptic. Despite the considerable influence from Greek-language and Hellenic textual culture on Egyptian textual culture,⁶⁴ especially during the Roman Period, the most parsimonious explanation is the former, not least when the evidence of contemporary Egyptian-language magical texts utilizing figures is considered.

18The spectrum extending from entirely “textual” to entirely “figural” amulets, laid-out above, is attested in ritual texts from Pharaonic and Graeco-Roman Egypt that were utilised for the benefit of a deceased (“afterlife texts”)⁶⁵ and those for the benefit of a living individual (“magical texts”).⁶⁶ For example, Book of the Dead recipes could be inscribed in a purely textual format, in a format integrating textual and figural content with embedded or superscript vignettes, or – in some cases – only as vignettes.⁶⁷ While Greek and Latin textual cultures preserve secondary sources⁶⁸ for contemporary periods which often describe, discuss, and even evaluate contemporary or past phenomena, such sources are not attested in Egypt. Hence, one must reconstruct a model from the primary sources with which to inform how the ritual instructions describe the production and function of those ritual texts. Through these models, scholars are then able to elucidate how the invocations of ritual texts apply efficacy to them, their accompanying figures,⁶⁹ or how figural representations were conceptualised⁷⁰ as inherently efficacious.⁷¹

19This overlap between funerary and magical figures, and their inscription as part of short compositions or “funerary amulets” requires more close consideration, not least because the tradition of inscribing afterlife texts with and without figures persisted until the very latest attested corpus in Demotic,⁷² and thereby into a temporal frame (the 2nd century) contemporary with the Graeco-Egyptian magical papyri.⁷³ After all, there is demonstrable evidence of transmission between the figural and textual culture of afterlife and magical texts in different stages of the Egyptian language in (at least) the Theban region during the 1st to 3rd centuries, and so it may have been in this period that the utilisation of figural culture previously restricted to Egyptian textual culture was transmitted into other forms. This would thus have been, most notably, in Greek-language magical texts and magical gems intended for a broader clientele, serving as a springboard for the transmission of textual and figural magical texts into the Late Antique and Islamic Periods, as well as beyond Egypt.

Inscribing figures in the Coptic magical text tradition

20So far I have hypothesised that Crum xxiii_2_1 belongs to a tradition of figural amulets/deposits which could have inherent efficacy (due to their form), or efficacy applied to them (through accompanying ritual practices). It is now necessary to contextualise its production in the Coptic magical text tradition, which preserves ritual instructions for the production of media inscribed with figures, and ritual practices directly associated with them.

21Many Coptic magical texts include ritual instructions for the production of amulets bound to⁷⁴ or worn⁷⁵ by the client, with the majority of these being to inscribe invocations, voces magicae – “name(s)”⁷⁶ and “power(s)”⁷⁷ – or a composition referred to as an amulet⁷⁸ onto a clean papyrus,⁷⁹ a clay vessel,⁸⁰ Aswan bowl,⁸¹ white palm fibre,⁸² or even an alabaster tablet,⁸³ as well as ostraca.⁸⁴ In other cases, the inscribed material is to be buried (ⲧⲱⲙⲥ) or cast (ⲛⲟⲩϫⲉ) somewhere specified. Most relevant with respect to Crum xxiii_2_1 are the cases in which the practitioner is instructed to inscribe figures.⁸⁵ Material which is figural and textual by nature can be produced in beneficent magical practices: the recipe of P. Heid. Inv. Kopt. 685 (TM 102074; c. 951-1000 CE) 10/1-18, which is accompanied by an avian-headed anthropomorphic figure labelled as the “young swallow” (ⲙⲁⲥⲃⲏⲛ), specifies that “in the moment that NN wears (ⲫⲱⲣⲓ < φέρειν) your figure (ⲥ̅ⲩ̅ < ζῴδιον)” (10/8-9), i.e. the produced amulet is worn by the client, “you (i.e. the figure) will watch over him all the days of his life against every evil spirit and unclean spirit” (l. 9-17) – although there are no ritual instructions telling us which specific medium the figure should be copied onto, nor where it should be worn.⁸⁶ Another recipe on p. 15/19-23 invokes “Jesus Christ, [to] help NN and everyone who will wear (ⲫⲱⲣⲓ < φέρειν) the nine guardians, whose names are… ”, with the names of those guardians roughly matching those of the figures inscribed on p. 12. A similar practice, but for healing rather than protection, is featured on p. 18/1-14, on which the names, powers, amulets, and dwelling places of Sabaōth are invoked to quench the fire of the client’s fever, as was done for the Three Hebrews in the fiery furnace of Nebuchadnezzar in the Book of Daniel, with voces magicae and figures depicting those Three Hebrews following.

22However, material which is both figural and textual by nature can also be produced in malign magical practices: the second recipe of P. Heid. Inv. Kopt. 681 (TM 99609; c. 967 CE; l. 28-52) is a curse to make a woman hated – the first recipe being for the opposite outcome. In this practice, the practitioner is instructed to write names and figures (ⲥ̅ⲱ̅ϯ̅ⲟ̅ⲩ̅ from ζῴδιον) as well as amulets (ⲫ̅ⲩ̅ from φυλακτήριον) on a potsherd (ⲡⲃⲓⲧ ⲡϭⲁⲗⲉϩ) with menstrual blood (ⲡϣⲣⲱ) which is then to be fired and placed at a crossroad. In addition, the erotic recipes of P. Heid. Inv. Kopt. 500 + 501 (TM 102087, c. 701-900 CE), in which the ritual instructions are in Arabic and the invocations in Coptic, feature recipes in which voces magicae and figures are to be inscribed upon an Aswan bowl with menstrual blood and hidden at the door of the target⁸⁷, in an empty well,⁸⁸ or are to be written upon a parchment,⁸⁹ a shroud,⁹⁰ a parchment with dove’s blood to be bound to the upper arm of the client, or buried where the target passes by.⁹¹ Another separation recipe is P. Heid. Inv. Kopt. 682 (TM 99576, 967 CE), which features the ritual instructions to write the figure (ⲥⲱ̅ϯ̅ⲟ̅ⲩ̅ⲛ̅ from ζῴδιον) on paper (ⲩⲁⲣⲏⲕ from ورقة.) and burn it in order to “undo” the couple.

23Despite this summary of numerous examples, there are no formularies which attest to a collection of figures, charaktēres, and voces magicae which are directly comparable to Crum xxiii_2_1. Furthermore, only a collection of malign magical practices (separation and erotic recipes) from a bilingual (Arabic-Coptic) handbook, in which the ritual instructions are written exclusively in Arabic and the invocations in Coptic, specify explicitly that the figural products are to be inscribed onto a parchment.⁹² Hence, with ritual practices for personal protection, healing, lustful desires, and curses of hatred and separation all requiring the production of inscribed media to be worn or deposited in order to bring about their desired outcomes, it is also not possible to narrow down the hypotheses regarding the connotations implied by the tableau of Crum xxiii_2_1 based on these parallels alone.

Coptic figural culture, in magical and canonical texts as well as art and iconography

24While the textual sources relating to the production of figural amulets or textual amulets with figures do not provide clear conclusions as to the function of Crum xxiii_2_1, it may be possible to infer its use by investigating more closely its figural features: where they may have come from and what their connotations may have been. In the Coptic and Islamic Period, figural culture was by no means restricted to magical texts; comparable figures can be found among tableaux within monastic and ecclesiastical milieus, in the form of – among others – illuminated manuscripts as well as wall paintings, and thus upon media which would have been familiar to both literate and illiterate viewers.

25The charaktēres, whether letter- or otherwise derived, are common to the majority of magical practices which require the inscription of another medium, but certain morphologies could also – especially in the centuries after their initial emergence – have been understood as adaptations or reinterpretations of iconography known from elsewhere. For example, the four-pointed cross (crux immissa: †) which terminates in rings and is comparable to a ⲑ-derived charaktēr, is also found ornamenting eulogia loaves⁹³ and the clothes of icons in churches,⁹⁴ and – in its most elaborate form – is instantly recognisable as a Coptic cross complete with ornate tassels.⁹⁵ By comparison, six-pointed stars (monogramma christi: ⳩) which also terminates in rings, are known in Coptic iconography through a pattern of dots in the equivalent shape, especially in the adornment of clothing⁹⁶. The eight-pointed star is already ubiquitous in the figural culture of the magical tradition through magical gems from the Late Antique⁹⁷.

26There is also the cross formed from a rhombus terminating in four equilateral triangles found under the rightmost ‘owl’ of Crum xxiii_2_1, and parallels to this morphology have been identified by Horak in various manifestations of Coptic Christian figural culture,⁹⁸ while they also appear in magical texts.⁹⁹ However, the significance of this rhomboid is most clearly demonstrated when compared with a similar attestation in recipe two (section 3) of P. Heid. Inv. Kopt. 500 + 501. Between two figures sits not only an eight-pointed star like that found on Crum xxiii_2_1, but a rhomboid terminating in four equilateral triangles above a triangle, one point of which terminates in a ring,¹⁰⁰ not unlike that to the right-hand of the rhombus on Crum xxiii_2_1. The figures and their accompanying charaktēres and voces magicae are to be written with menstrual blood on parchment before being buried somewhere, while the invocations that follow are to “give peace and desire quickly and with haste through the power of this other dozen great names” (l. 36-38). The implication that these figures and their accompanying charaktēres and voces magicae would be inscribed onto a figural amulet and deposited in order to bring two parties together – as that very tableau depicts through its figures, bound by an eight-pointed star, rhombus, and triangle –, has wide-reaching implications for the analysis of the function of Crum xxiii_2_1, to be revisited below.

27The L-shaped avian figure, by comparison, is not dissimilar from those which feature in tableaux depicted on magical gems¹⁰¹ but also in illustrated Coptic manuscripts,¹⁰² and – of course – magical texts.¹⁰³ Notable is the fact that an aforementioned manuscript (n. 102) features at least seven different morphologies¹⁰⁴ for the same marginal ornament of a dove/sparrow across its 65 leaves. The ‘owls’, however, when their anthropomorphic faces are coupled with the wings of the rightmost’s and the clothing of the leftmost’s, could be interpreted more earnestly through similarities with depictions of angels in Coptic iconography,¹⁰⁵ and those in magical texts.¹⁰⁶ Although feet appear absent in the left-most and perhaps stylised in the right-most, the variation between arms and wings could, in fact, reflect either two representations of the same angel, i.e. on land and in flight, or of two angels undertaking different functions. The leftmost’s ‘banjo’ can be compared to the staff wielded by the “young swallow” depicted on page 10 of P. Heid. Inv. Kopt. 685, or the depiction of Daveithea in MS Or. 6794 (TM 100017) described in the recipe itself as “the one who has the golden bell (ⲡⲉϣⲕⲗⲕⲓ̈ⲗ ⲛ̅ⲛⲟⲩϥ) in his right hand and the spirit lyre (ⲧⲕⲓ̈ⲑⲁⲣⲁ ⲙ̅ⲡⲛ̅ⲁ) in his left hand” (l. 7-8). This does, however, make more sense in the context of that recipe, intended to give a good singing voice.

28The unmistakable Apa Baula, however, can be compared with greater ease in all of his facets to features in figural culture of multiple sources. His hair or headdress is reminiscent of the plumes that completed the headgear of anthropomorphic and composite figures in the vignettes of afterlife texts,¹⁰⁷ but has been referred to in the context of Coptic illuminations as a “Strahlenkrone”, according to a parallel which depicts an angel wearing such a “crown”.¹⁰⁸ However, given that the auspicious seven strands of hair/plumes of a headdress or spikes of a crown are also found on a figure which depicts the confounded target of the malign practice of P. Heid. Inv. Kopt. 679 (recto and verso), perhaps this was simply a stylised way of depicting hair, or someone engaged in practicing, being invoked within, or being the target of, ritual practices. Certainly, the figure of Christ in P. Heid. Inv. Kopt. 500 + 501 has different hair, as do the anthropomorphs depicted on Cairo Coptic Museum 4959 (TM 100008) and 4660. By contrast, Apa Baula’s facial features follow the typical conventions for depicting human faces in Christian Egyptian figural culture, whether in magical texts¹⁰⁹ – as in the preceding Graeco-Egyptian tradition¹¹⁰ – or in illustrated manuscripts.¹¹¹ The morphology of his body and the lines presumably depicting his clothes also resembles those in other magical texts,¹¹² as do his unmistakably anthropomorphic arms¹¹³ and hands.¹¹⁴

29The Destroyer, ⲡⲣⲉⲃⲧⲁⲕⲟ ⲥⲁⲣⲁⲧⲁⲗⲏ, can likewise be interpreted by reference to the specific morphological features utilised to construct this extraordinary zoomorphic composite. From the top-down, the preceding subjective description can be significantly refined with reference to specific parallels. The head is similar to that atop the anthropomorphic body of EA 10414¹¹⁵ (TM 99562; 6th/7th century; fig. 5), part of a tableau on the recto underneath two sections featuring erotic recipes. Although the ritual instructions on the verso to “write the figure at the bottom of the new pot” (verso l. 18-19) may relate to a recipe for “favour and blessing and love” (verso l. 11-14), also on the verso, it is also possible that either these ritual instructions refer to figures beneath the text which were broken off, or that they apply to the recto text as well. Similarly, it is equally possible that ritual instructions for inscribing this tableau are lost beneath the tableau itself on the recto. Nonetheless, these two recipes are both intended to bring a woman to the client, and thereby fit with such malign figures. However, with other aspects of the head’s morphology, including the hair, teeth, and eyes, being different, one cannot draw too much from this ‘parallel’.

30The closest zoomorphic morphology I have been able to identify, by comparison, is that of the lions accompanying Paul the Hermit as depicted in the northern nave of the Monastery of St. Paul (fig. 6). ¹¹⁶There, not only are the pointed ears identical, the straight horizontal eyebrows, ridge of the nose, and adjacent circular eyes are depicted similarly, with the only significant difference being that the termination of the nostrils of the lion’s nose is not depicted on The Destroyer, because the snout/muzzle serves this purpose. In addition, while the bodies of the lions are in profile, the faces are front-facing, as with The Destroyer – common to depictions of mammalian quadrupeds.¹¹⁷

31As can be seen by comparison with the “camel-monsters of Menas” depicted on the “Dome of Martyrs” in the Monastery of St. Paul (fig. 7),¹¹⁸ as well as with the monster being speared by Curiacus in a mural on the northern nave (fig. 8)¹¹⁹, teeth were commonly depicted in Egyptian figural culture of the Late Antique period not within an open mouth,¹²⁰ but over a closed snout/muzzle. Hence, the lines inwardly perpendicular to the outer frame of the snout/muzzle of The Destroyer without a doubt depict its multiple teeth.

32The Destroyer’s body, at once familiar and fantastical, finds no parallels among the figures of Coptic magical texts,¹²¹ but the curling-up tail is found in depictions of animals of attack, for example lions¹²² (ⲗⲁⲃⲟⲓ) and foxes¹²³ (ⲁⲗⲥⲱⲗⲱⲩⲕⲓ), in illuminated Coptic manuscripts, as well as in those of dogs¹²⁴ in wall paintings. Hence, the resemblance of The Destroyer to that of a horse or camel can only be true in form, but not in terms of its understood nature, which would be closer to a wild carnivore. In addition, given that the tail terminates in a fork and is bisected perpendicularly with shorter lines terminating in rings, one gets the impression that the intention was to represent spines jutting out from the tail, in a manner comparable to the way the teeth are depicted. However, depictions of hairy tails terminating in a fork¹²⁵ suggest that The Destroyer’s tail could also be an elaboration of a much more standardised depiction in Coptic figural culture.

33While the “hump” could be compared with that of the aforementioned “camel-monsters of Menas”, or those depicted in manuscript illuminations,¹²⁶ it also displays similarities to depictions of horses carrying loads.¹²⁷ Yet, the fact that the face and tail are not those of either a horse or camel leaves the feature on The Destroyer’s back somewhat of an enigma.¹²⁸ What is clearer, however, from the “camel-monsters of Menas”, and elsewhere,¹²⁹ is that in comparable figural culture the most common way of depicting the legs of horses or camels is as bent in the middle and terminating in rounded forms representing hooves.¹³⁰ From this survey, The Destroyer seems therefore to be a composite of multiple elements – some of which have been identified – yielding a long- and tooth-snouted hooved quadruped, albeit with nine legs, exhibiting a protrusion resembling the hump of a camel and the curling tail of an attacking carnivore.

“Figural” from “textual authority”? Which Apa “Paul” and who was ⲡⲣⲉⲃⲧⲁⲕⲟ ⲥⲁⲣⲁⲧⲁⲗⲏ?

34With parallels in the figures of Demotic, Greek, and Coptic magical texts not forthcoming, and the implications from the wider figural culture of Coptic and Islamic Egypt limited, the final route of investigation is to try to identify the mythical precedent to which the tableau of Crum xxiii_2_1 alludes – and thereby whose efficacy it manifests in figural form. Following this, it will then be possible to examine how the primary protagonists in the tableau, Apa Baula and The Destroyer, inform us about the desired outcome of producing and wearing such a figural amulet or depositing such a ritual deposit.

35The name Paul could refer to a number of figures in Church and Coptic history, but it is in the figure of Saint Paul of Tamma,¹³¹ whose Vita is preserved in both Coptic¹³² and Arabic¹³³ and tells of a monk living supposedly at the end of the 4th and into the 5th century in Middle Egypt, that the most likely candidate is found. When Paul was 18, he withdrew as a hermit to the mountain of Touho, where he was instructed by the monk Hyperichus. 54 years later he was joined by Ezekiel, whom he instructed, and it was with Ezekiel, after various events with other monks and apostles, and following the six deaths – seven in the Arabic version – and resurrections of Paul that a certain episode took place. This episode is, following convention, narrated by Ezekiel, who describes sitting alone in their dwelling place when a man came to the door and called to him asking whether he was Ezekiel, for, the visitor said, he had something to share with him (764.7-14). The visitor revealed that he had been seeking materials for a bath with some other men and they had been robbed, left only with their lives (764.14-765.3). When they ran away from the robbers, they stumbled upon a “large man who was bound and had been cast into a valley”, who had told the visitor to locate the dwelling place of Ezekiel because the bound man was his father: Paul (765.4-9). Taken in by this reference to his father, Ezekiel endeavoured to find him, and was led deeper into the desert by the visitor for three days (765.9-766.1).

36As they were travelling, the visitor underwent an unexpected metamorphosis. Ezekiel narrates: “But as he (the visitor) was speaking with me, he changed. He became a great Ethiopian, his eyes filled with blood and his whole body full of spines, smelling badly like male goats. He said to me: “Did you not recognise me; Ezekiel?” I said to him: “No!”. He said to me: “It is I who once made the well (πηγή) waterless. Your father struck me, until I filled it as it was before. I have suffered, enduring you and your father. (Neither of) you have become wise.” I said to him: “Then you are the Devil! May the Lord punish you for all the evil things you have done to the servants for Christ!” Then he leaped after me like a wild (ἄγριος) young lion, wishing to kill me. But I cried out: “My father, help (βοηθεῖν) me!” Suddenly, my father heard my voice as if he was near me and, immediately, he arose and came towards me through God. When my father approached him, the Devil changed. Immediately, he took the form of a monk, wearing animal skin and small bundles of palm-leaves. He turned to my father and bowed before his feet according to the custom of the monastic brethren. Immediately, my father made a mark around him. He could not move to that side or this. That is how he knew it was the Devil. And he restrained him, bound him by his hands and feet and rolled him down into a valley. We left him and walked (away)”.¹³⁴

37This episode is followed by an epilogue in which Ezekiel asks Paul from where he acquired the ability to tackle the Devil in such a way. Paul answers: “God gave it to me, so that I (could) punish (παιδεύειν) him according to my desire, for he defiled and he tempted (πειράζειν) us”.¹³⁵ After returning to their dwelling place (767.4-5), Paul undertakes particularly extreme ascetic practices, with Ezekiel describing how “he spent 40 days and 40 nights (thus): He did not eat nor did he drink, sitting upon a brick inside the dwelling-place looking at a mirror in a window. He did not shut his eyes during these 40 days until they burst and bled upon the ground”.¹³⁶ Following these practices, “the holy archangel Michael came from heaven in the hour of the light of the Lord’s Day (κυριακός) after the completion of the 40 days. He sealed him and released him from all his sufferings. His eyes were set right in their way. Michael returned to the heavens in glory.”¹³⁷ And on that Lord’s Day, Paul and Ezekiel joined the monks and ate together, before departing south for another adventure in Siout (767.12-14).

38On balance, although the zoomorphic composite depicted on Crum xxiii_2_1 is not, admittedly, described in Paul’s Vita, from this episode one can establish a tangible precedent in the mythology of the Saint that could have inspired a tableau utilised in magical practice. The story establishes Paul’s God-given, and therefore legitimate, ritual power which he may use to encircle, bind, and tackle the Devil, no matter how savagely animalistic his form. The semantic range of the Devil’s description in this episode is notable, with the constructions “like male goats” (ϭⲓⲏ ⲛⲃⲁⲁⲙⲡⲉ) and “like a wild young lion” (ⲟⲩⲙⲟⲩⲓ ⲙⲟⲩⲓ ⲛⲁⲕⲣⲓⲟⲛ) both being metaphors relating to wild animals, the former recalling the morphology of The Destroyer in general terms, the latter the morphology of his head and the attacking pose of his curled-up tail, while the description of the Devil’s “whole body [being] full of spines” recalls the protruding spines of the tail. Thus, the description of this incarnation of the Devil draws directly from animalistic imagery, in direct contrast to the clothed communicators that the monks present themselves as. In addition to this animalistic imagery, the reference to the devil as an Ethiopian, and therefore black-skinned, relates directly to wider conceptions within (Egyptian) Christianity. As treated by Brakke, monastic literature is full of references to demons who would manifest before monks who are described as having black skin – while one effect of such stories is to “demonise Ethiopians”, another is to “Ethiopianise demons”.¹³⁸ Not restricted to Egyptian Christianity, the “black-skinned demon of early Christian monastic literature was born from the association of blackness with sin and evil that was common in numerous varieties of ancient Mediterranean religious discourse” – however, “it became Ethiopian (and not merely black) in Egypt”.¹³⁹ Given that manifestations of demons as animals was perhaps the most prevalent trope of monastic literature, this example appears to intensify the dread narrated by combining the two tropes.

39Returning to the tableau of Crum xxiii_2_1, the lines connecting Apa Baula’s arms to The Destroyer’s neck and body imply the acts of binding, trussing, and by clear analogy, total ritual control. When a comparison between the eight-pointed star, rhombuses, and equilateral triangle on Crum xxiii_2_1 are made with similar examples that are depicted between the twin figures of the aforementioned binding recipe of P. Heid. Inv. Kopt. 500 + 501, the connotations of binding and thereby control are borne out by both textual and figural parallels to Crum xxiii_2_1. What is more, the episode in the epilogue describing the Archangel Michael’s descent from and reascent into heaven may provide an explanation for the depiction of two angels, one in a more anthropomorphic, and one in a more avian form – the former also wielding a sceptre which has been compared above to a seal. Could one therefore see the squares, most of which feature charaktēres as part of their morphologies, as figural representations of the “mark/boundary” (ϣⲱⲗϩ) which Paul laid around the Devil in order to subdue him? With these hypotheses forthcoming, one important final consideration is the identity of the second primary protagonist of Crum xxiii_2_1, implied by its name: The Destroyer Saratalē.

40If Saratalē can be understood as a form of Sanataēl (see n. 39), a name of the Devil, it may be relevant to consider the appearance of this name in two Coptic magical texts. The first is the 10th century parchment codex P. Heid. inv. Kopt. 686 (TM 100022), a compilation of invocations for the laudation of the Archangel Michael. In this text, Sanataēl (ⲥ̅ⲁ̅ⲛ̅ⲁ̅ⲧ̅ⲁ̅ⲏ̅ⲗ̅) is given the epithet “the first creature” (ἀρχίπλασμα; 3/2-3/3, Kropp l. 37-38), a term frequently used to refer to the devil in Coptic.¹⁴⁰ The recipe in the bottom register of the 11th century parchment P. Heid. inv. Kopt. 683 (TM 100000) is, by contrast, a malign magical practice, and one which also features a four-figure tableau. The recipe itself invokes one Salathiēl (ⲥⲁⲗⲁⲑⲓⲏⲗ) “to go to NN and shatter her mind until she gets to her feet, comes to the house of NN and gives his masculine-desire to her female-desire through a shameless face” (l. 25-29). Thus, this practice is an erotic recipe to bring a target to the (practitioner-)client, and one which appears to invoke a lambdacised Sarataēl, with the orthographic variant ⲑ replacing ⲧ. Notable in light of the “Ethiopanised demons” in the Vita of Paul, discussed above, is the fact that Salathiēl in this text is given the epithet “the Ethiopian” (ⲡⲉϭⲱϣ, l. 20), making an association between Salathiēl and the Devil clear, and the equivalence of Salathiēl, Saratalē and Sanataēl very likely. Thus, when Ezekiel referred to the figure as “a great Ethiopian” in the Vita of Paul, the figure of Saratalē/Sanataēl, the Devil, would likely have been conjured in the minds of the audience.

41However, while Saratael is likely a manifestation of the Devil in the tradition from which these texts derive, there is also an archangel with the same name who appears in a Coptic doxology which forms part of the service of the evening and morning offering of incense.¹⁴¹ In a modern Bohairic version, this archangel is one of the seven archangels which praise as they stand before the Pantocrator and serve the hidden mystery. The Trinity is symbolised by Michael, Gabriel, and Raphael, whereas the final four are Suriel, Sedakiel, Sarathiel (ⲥⲁⲣⲁⲑⲓⲏⲗ)¹⁴², and Ananiel, who are the great holy luminaries entreating him (i.e. God) before creation.¹⁴³ Thus, although it is most unlikely that “the Ethiopian” or “The Destroyer” were conflated with this archangel, there is clearly more to this name than the particular associations treated here, and the names of the two angels – the heavenly luminary and the fallen Devil –, may have influenced one another.

42Having identified Saratalē with Sanataēl, “the Ethiopian” Salathiēl, and hence with the Devil in Coptic tradition, the figure’s epithet, The Destroyer, can now be considered. In Revelation 9:1-10, the fifth angel is described as blowing his trumpet, at which point a star falls from heaven and a key to the shaft of the bottomless pit is produced, which – when unlocked – unleashes monstrous locusts¹⁴⁴ who go about tormenting those without the seal of God upon them for five months. Their king is the angel of the bottomless pit, Abaddon (אֲבַדּוֹן) or “Destruction” in Hebrew, glossed as Apollyon (Ἀπολλύων) in Greek. However, in the Coptic New Testament the Hebrew “Destruction” ¹⁴⁵is referred to by a Coptic translation of the Greek: S ⲡⲉⲧⲧⲁⲕⲟ “The Destroyer”. Although grammatically not identical, there is no semantic difference between ⲡⲉⲧⲧⲁⲕⲟ and ⲡⲣⲉϥⲧⲁⲕⲟ, and this wreaker of carnage is also that which killed the first-borns in Exodus 12:23 and the grumblers in the warning against idolatry of Corinthians 10:10. Hence, if The Destroyer of Crum xxiii_2_1 may be related to the Coptic understanding of Abaddon, one should consider how Abaddon was conceptualised elsewhere by Egyptian Christians.

43In a treatment of Christian Paintings from Tebtunis, Walters¹⁴⁶ noted “a building containing Christian paintings” of the mid-10th century which “almost certainly formed part of a monastic complex”, including a study of the depictions of “The Punishment of Sinners”.¹⁴⁷ Central to this tableau was “a gigantic winged figure”, “Lord Abbatōn the angel of unrespecting death”,¹⁴⁸ whom Walters describes as having a snake entwined around its waist, anthropomorphic hands but claw-like feet, holding a rope from which hang the heads, torsos, and bodies of sinners. Elsewhere in the tableau Abdimelech, “the angel of punishments”, is depicted, as well as “the decan chewing the soul”, and a donkey-headed demon “who serves the Lord of Darkness”. If such a fantastic representation of torment was once extant in a church in the Fayum, providing figural representations of descriptions preserved in the textual culture of the Apocalypse of Paul and the Discourse on Abbadon, the cross-fertilisation and transmission between figural and textual cultures should not only be emphasised, it should be considered that the apparently fantastic figures appearing in magical texts are actually not quite as remarkable as they may at first seem to modern observers.

44In the Discourse on Abbadon by Timothy, Archbishop of Alexandria (MS Or. 7025; 982 CE),¹⁴⁹ the author describes going to Jerusalem in order to discover “the creation of Abbadon” by searching through the books in its library which had been produced by the Holy Fathers and Apostles and deposited there. Subsequently, Timothy narrates an account of creation in which the angel Mouriēl reports to God how the man created in his image has transgressed his command by eating from the tree. As a result, Mouriēl is transformed into Abbatōn the Angel of Death (ⲁⲃⲃⲁⲧⲱⲛ ⲡⲁⲅⲅⲉⲗⲟⲥ ⲙ̅ⲙⲟⲩ), to be “terror in everyone’s mouth” (241.7-8) and “whose form (ⲉⲓ̈ⲛⲉ) and image (ϩⲓ̈ⲕⲱⲛ) will be complaint, wrath, and threat in all souls” (241.9-11). The description of the morphology of the Angel of Death, with “eyes and face… like a wheel of fire”, “the sound of [his] nostrils… like the sound of a lake of fire”, etc. (241.12-28), certainly has little to do with the bestial depiction of The Destroyer on Crum xxiii_2_1. Rather, these descriptions instead relate him to the geography of hell in Coptic Christian conceptions – wheels and lakes of fire.¹⁵⁰ However, the composite and monstrously large description of this demonic being is notable because it attests an instance of a textual description becoming a reality in the figural culture of at least one church in the Fayum.

45This example may be complemented by a few of the many instances of theriomorphic depictions of demons in Coptic literature. The demons depicted alongside the Angel of Death in the tableau from Tebtunis find a textual parallel in the description of the Punishment of the Sinners in the Apocalypse of Paul (MS Or. 7022),¹⁵¹ another instance of textual culture transcending the written word and becoming manifest in figural form. The “Powers of Darkness” (ⲛⲉⲝⲟⲩⲥⲓⲁ ⲙ̅ⲡⲕⲁⲕⲉ) are described as having the faces of animals (lions, bulls, bears, serpents, ravens, ibises, donkeys, crocodiles, and “beasts”) with smoking mouths, horns, eyes, or tongues of fire, teeth of iron, iron or black breastplates of fire, and wielding swords, iron, saws, knives, spears, forks, and blades (556.33-557.12). Likewise, the Apocalypse of Elijah,¹⁵² describes innumerable angels whose faces are like leopards (πάρδαλις), whose teeth protrude from their mouths like pigs, whose eyes are clouded with blood, whose hair is like women’s, and who have flaming whips in their hands (4/15-5/4), while a “great angel” is described later in a similar manner, with hair was spread out like a lion’s or a woman’s, his teeth protruding like a bear (ἄρκος), and his body like a snake which desired to swallow the protagonist (8/2-14). These composites, attested throughout apocryphal literature, exhibit a veritable menagerie of zoomorphic composites, preserved and transmitted in the Coptic-language tradition, of which Zandee has summarised some of the most diverse examples.¹⁵³ In short, these episodes feature and – importantly – describe a veritable menagerie of zoomorphic composites which would outclass many of the most monstrous figures extant on Coptic magical texts, suggesting also notable influence between apocryphal and magical textual and thereby figural traditions.¹⁵⁴

46Such monstrous manifestations aside, there is a final point of reference relevant to nuancing the identification of The Destroyer. In Shenoute’s writings, such as in the opening section of letters of Canon 1,¹⁵⁵ in which he describes how God will save souls against the “sword of the enemy” and “make them whole in his kingdom which is heaven”, he tells how “The Destroyer came into you (pl.). The Destroyer mastered/ruled a share of your (pl.) hearts. And he took it captive (αἰχμάλωτος)” (195.24)^.156 Later, Shenoute describes how The Destroyer has overturned the wall of the enclosure of the assembly and destroyed all the botanical life which was in it (196.1-11) so that all that remains in the brothers is the manner of Sodom and Gomorrah. Following these sexual connotations is the punishment inflicted by The Destroyer on promiscuous men, whose souls will fill with “the demonic spirit” (ⲡⲛ̅ⲁ̅ ⲛ̅ⲇⲓⲁⲃⲟⲗⲟⲥ; 211.5-20), while the Destroyer destroys their seed. Also significant in this section is that The Destroyer is equated directly with Satan (211.9), demonstrating that the epithet ⲡⲣⲉⲃⲧⲁⲕⲟ was associated not only with the manifestation of evil that was Abbadon (as in the biblical tradition), but also with the Devil himself.¹⁵⁷ Thus, this consideration relates the analysis of “The Destroyer Saratalē” with the episode from the Vita of Saint Paul of Tamma, in which he successfully tackles the Devil in a monstrous and anthropomorphic form. When these examples are then taken into consideration alongside the interpretation of the name ⲥⲁⲣⲁⲧⲁⲗⲏ as Satanael, the Devil, the identification of Apa Baula’s adversary as ‘the adversary’ seems robust.

Inscribing and invoking zoomorphic composites – tradition and transmission

47This study has treated how zoomorphic composites, as described in textual and depicted figural culture, can inform upon even the most unparalleled examples of figures from the tradition of Coptic magic. At the outset, the Egyptian tradition of tableaux of figures (vignettes) in ritual texts – whether afterlife or magical – and of those zoomorphic composites depicting both deities and demons was brought into focus. In broad terms, composite zoomorphic manifestations were conceptualised as exhibiting divine or demonic efficacy in Egypt for millennia:¹⁵⁸ whether in the 13th/12th century BCE hieratic recipes against the shꜣḳḳ demon, who is described in grotesque form as a faeces-eating demon with eyes in back of its head, a tongue in its anus, its (mammalian) paws contorted over one another,¹⁵⁹ and depicted with a mammalian tail,¹⁶⁰ or in apocryphal Christian texts preserved in Coptic where the powers of darkness are described as animal-headed composites who took delight in tormenting humans. Yet this was not simply the case in magical texts, but also in narratives about demons such as those found in monastic literature, where one of the foremost features of the manifestations of the devil or demonic entities is their zoomorphic nature, whether visible and visual, such as the zoomorphic ‘pagan’ idols described in the Vita of Severus by Paphnutius in the Histories of the Monks of Egypt, or invisible and behavioural or inherent, such as the description of demons who make the earth to shake due to their galloping hooves,¹⁶¹ described in the Vita of Apa Moses.

48Consequently, it has been seen in the Vita of Paul of Tamma that it was – to borrow from Brakke – not only certain animals which were demonised, but also certain demons which were animalised. To animalise required an emphasis not exclusively on appearance, but also on behaviour and intrinsic qualities. In the Vita of Paul of Tamma, lions are conceptualised as savage, and goats as exuding an ungodly odour – behavioural and sensory qualities respectively. Yet these were facets which could not be so easily depicted in figural culture through static, unannotated figures. In the text, the demonic associations of the black-skinned Ethiopian are coupled with descriptions which specifically highlight animalistic behaviour, rather than simply an animal-like visual manifestation. When approaching the interpretation of zoomorphic figures in Coptic magical texts, therefore, we must consider that the visual elements being depicted cue the conceptualisation not only of physical features, but other behavioural and sensory qualities particular to the animals whose bodies are depicted. Might not a visual reference to the forepart of a lion, for example, not only cue the associations of its muscular frame, but also the savage nature described in the Vita? Such considerations highlight how what appear to be relatively simple descriptions in textual, and depictions in figural, culture of zoomorphic elements may instead need to be set within a wider interpretative frame, which takes into account the primary conceptions held by contemporary communities about the most salient aspects, whether physical, behavioural, or intrinsic, of those animals in question.

49Returning at last to the tableau of Crum xxiii_2_1, it is clear from the preceding considerations that any conclusions regarding such a complex composite are likely to miss notable elements of the interpretative framework that may once have been cued by the depiction of The Destroyer. However, what is illustrative is that from a consideration of the figural culture attested in Coptic magical texts, as well as in illuminated manuscripts, and – most notably – murals from churches and monasteries, the composite figures which constitute its primary protagonists may not have been considered quite as fantastical by contemporary observers as they are by modern observers. Although the “Christian paintings from Tebtunis” are one isolated and currently exceptional attestation of a public manifestation of the darkest elements of Coptic figural culture, they may have been representative of a broader phenomenon. If so, the veritable menagerie of demonic beings that were depicted and thereby present in figural culture, when coupled with those that were described and thereby heard of or read about in textual culture, would have permeated contemporary milieus with a notion of both a lived sphere and an otherworldly sphere populated by zoomorphic composites.

50In addition, if – given his paradigmatic status in the Coptic church – the sermons of Shenoute can be considered as a precedent from which to draw, the manifestation of the Devil, described as the adversary of so many Coptic monks, as The Destroyer (ⲡⲉⲧⲧⲁⲕⲟ), who wreaks havoc on both the faithful and sinners, would have been a recognisable epithet to a contemporary audience. Thus, a depiction of The Destroyer could have drawn upon the figural features of his namesake Abaddon the Destroyer, known from other sources of textual, and thereby figural, culture.¹⁶² Furthermore, if Saratalē (ⲥⲁⲣⲁⲧⲁⲗⲏ) is to be understood as a rendering of Sanataēl, and thus Satanaēl, one reference to a manifestation of the Devil from Coptic textual culture is followed by another in the tableau of Crum xxiii_2_1. With such an adversary, it is unsurprising that God imbued his saints with the ability to tackle and overcome his Adversary, whenever and wherever he was encountered.

51Thus, the episode from the Vita of Paul may well have served as the mythical precedent through which practitioners stemming from, and practicing within, a milieu steeped in Coptic figural and textual culture would either have been able to bind the target of Crum xxiii_2_1 through the example of Paul binding his Adversary or instead protect a client wearing it from being bound, by visualising the efficacious Apa Baula who, with the blessing of God and the assistance of the Archangel Michael, overcame The Destroyer.