Faraone ————— Greek, Roman, and Byzantine Studies 62 (2022) 280–308 Article copyright held by the author(s) and made available under the Creative Commons Attribution License CC-BY https://creativecommons.org/licenses/by/4.0/ Notes on Magical Texts and Gems Christopher A. Faraone 1. Disambiguating two images on a green jasper gem On the obverse of a somewhat large (2.9 x 3.3 cm) green jasper once in the Louvre, but now lost (D&D 122), we see enclosed within a circle formed by a snake biting its own tail ( fig. 1.a)1 a unique and quite puzzling version of the scene of Anubis embalming Osiris, who lies face up on a lion-couch, which is in turn being fed by a female as it (the lion-couch) tramples upon a ——— Abbreviations: BM: S. Michel, Die magischen Gemmen im Britischen Museum I–II (London 2001) CbD: Campbell Bonner Magical Gems Database (http://classics.mfab.hu/talismans/object/list) D&D: A. Delatte and P. Derchain, Les intailles magiques gréco-égyptiennes de la Bibliothèque Nationale (Paris 1964) DT: A. Audollent, Tabellae Defixionum (Paris 1904) GEMF: C. A. Faraone and S. Torallas Tovar, The Greco-Egyptian Magical Formularies I (Berkeley 2022) GMPT: H. D. Betz, The Greek Magical Papyri in Translation (Chicago 1986) LIM: A. Mastrocinque, Les intailles magiques du Département des Monnaies, Médailles et Antiques (Paris 2014) LMPG: L. Muñoz Delgado, Léxico de magia y religión en los papiros mágicos griegos (Madrid 2001) SMA: C. Bonner, Studies in Magical Amulets (Ann Arbor 1950) 1 Figures 1.a-b are after the photo of the plaster cast of D&D 122, which has been flipped horizontally so that the inscriptions can be read in the proper direction. The gem itself seems to have disappeared; Attilio Mastrocinque per email informs me that when he was working on his re-edition of the corpus, it could not be found in the magazine and as a result it does not appear in LIM. For discussion see A. Nagy and K. Endreffy, “Apolyson apo pankakou!” A Recently-surfaced Gem,” in G. Rocca and G. Bevilacqua (eds.), Gift of a Book: Studi in memoria di David Jordan (Alessandria 2020) 179. CHRISTOPHER A. FARAONE 281 ————— Greek, Roman, and Byzantine Studies 62 (2022) 280–308 skeleton. Along the inside of the ouroboros we see signs of another problem, handbook instructions that have been mistakenly carved onto the stone: ο λέοντος γλύφε2 ἀναµέσον δυέων ἱεράκων βασιχε (“engrave the […] of the lion between the two falcons: BASICHE”). It seems, then, that the scribe was supposed to en- grave the word BASICHE between the two falcons, which face each other on either side of Anubis and, as we can see in fig. 1.a, these letters would indeed fit in this space, which is now taken up by the word ἀναµέσον. But instead, the scribe erroneously copied out the text of the instructions in a long curve over three- quarters of the gem. The meaning of ο λέοντος is not obvious at first glance, but since the omicron with overstroke, or a rectangle representing it, is a common way of abbreviating the word onoma in the magical formularies, it was probably misunderstood by the gem-engraver.3 The original handbook instructions must have read, therefore: “Inscribe the name of the lion between the two falcons: BASICHE.” This instruction, in fact, recalls those for a gold ring with a copper signet prescribed by Alexander of Tralles for bladder pain (II 475.18–23 Puschmann): “Make a signet- stone of Cyprian copper ore and carve it with a lion, a moon and a star and in a circle (κύκλῳ) inscribe the name of the beast (ὄνοµα θηρίου, i.e. the lion) … and wear it on your pinky or forefinger.” On the reverse of the gem ( fig. 1.b) we again find handbook instructions mistakenly copied out—ὀπίσω δὲ τοῦ λίθου κύκλῳ πολιν(?) τὰ ὀνόµατα ταῦτα (“on the back of the stone in a circle polin[?] these names”)—followed by six more powerful names: CHUCH BACHUCH BAKAXICHUCH BAKABEICHUCH BADITIPHO BAINCHOOOCH. Here, too, the instructions are a bit muddled. Delatte and Derchain correct πολιν to πολιῷ and take the word as a modifier of circle, e.g. “a gleaming circle,” but, when κύκλῳ 2 This is the suggestion of Attilio Mastrocinque. D&D read ταῦτα. The word is not so clear in the photograph, especially the fourth letter (the fifth is not visible). 3 This and the parallel given below from Alexander of Tralles are the suggestions of the anonymous reader, to whom I give thanks. 282 NOTES ON MAGICAL TEXTS AND GEMS ————— Greek, Roman, and Byzantine Studies 62 (2022) 280–308 appears without a preposition in such handbook instructions, it is always adverbial and never modified by another word.4 It is perhaps better to understand πολιν as a mistake for πάλιν. In any event, it seems clear that if the gem-cutter had correctly followed his instructions on the back of the gem, he would have inscribed the six magical words in a circle along the edge of the gem. What, then, was supposed to fill the empty space in the middle of the reverse side of the gem? The answer lies, I suggest, in the crowded and illogical image on the front of the stone, where the scribe (or his exemplar) apparently misunderstood the handbook recipe (for the third time), because he conflated two images that usually appear separately: (i) the scene with Anubis embalming the Osiris mummy on the lion couch (see fig. 2); and (ii) a differ- ent scene of a lion trampling a corpse (see fig. 3). It seems that the appearance of the lion in two different descriptions of the gem in the recipe confused the gem-cutter, whose eye skipped from the middle of one description to another presumably in the same column of the handbook. A similar mistake occurred on another lost gem, whose design was fortunately copied in the early seventeenth century by Nicolas-Claude Fabri de Peiresc, a friend of the painter Peter Paul Rubens. According to his sketch ( fig. 4), a lion-headed ouroboros serpent ran along the edge of the oval gem, and twelve lines, almost all of it Greek text, were inscribed within the ser- pentine border:5 4 See e.g. LMPG s.v., listing seven other examples—five of them from PGM IV—where κύκλῳ without any preposition is followed directly by τοὺς χαρακτῆρας τούτους (PGM III 293), τὸν λόγον τοῦτον (IV 2016, 2048, 2070), or τὸ ὄνοµα/τὰ ὀνόµατα (2138, 2693). Only in VII 468 do we find another word inserted in between: κύκλῳ αὐτοῦ τὰ ὀνόµατα. 5 R. D. Kotansky and J. Spier, “ ‘The ‘Horned Hunter’ on a Lost Gnostic Gem,” HThR 88 (1995) 315–337, whose text and translation I use; fig. 4 is after their illustration p.337. For other examples of such scribal confusion see D. R. Jordan, “Il testo greco di una gemma magica dall’Afghanistan(?),” in A. Mastrocinque (ed.), Atti dell’ incontro di studio: Gemme gnostiche e cultura ellenistica CHRISTOPHER A. FARAONE 283 ————— Greek, Roman, and Byzantine Studies 62 (2022) 280–308 ὁ Ἰοαηουαυη, ὁ Βακαξιχυχ, ὁ Κ- ερατάγρας, οὗτός ἐστιν ὁ πρωτοπάτ- ωρ, ὁ τοῦ σώµατος µ- όνος ὤν, διὰ πάντ- ων πορευ⟨ό⟩µενος. Σο- λοµῶντος {ι}σφραγί- ς ÚƵƵƵ (fig.) ὁ δρα- κὼν ἔστω λεοντοκέφ- αλος The Ioaêouauê, the Bakaxichuch, the Keratagras. This is the First-Father, the one being single of substance, passing through all (things). The6 Seal of Solomon ÚƵƵƵ (fig.). Let the snake be lion-headed. The space marked “(fig.)” refers to a tiny image of a serpent in line 9 of Peiresc’s drawing ( fig. 4) and the words that follow—“let the snake be lion-headed”—refer to this same image. As the editors suggest, both seem to have been copied mistakenly from a handbook recipe, which presumably recommended carving the familiar image of the lion-headed snake-god Chnoubis (with his characteristically single-looped tail) that we see in fig. 5. There were, moreover, some further problems with the text on this lost gemstone sketched by Rubens’ friend. We know that the ÚƵƵƵ symbols were a version of the Jewish tetragramaton—a kind of secret name of Jahweh that is often equated, as it is here, with (Bologna 2002) 61–68, for a gem in Moscow, whose text begins “This is the logos: Abaichôrmuid,” or LIM 286, where around the shield of the anguipede the scribe wrote the words “the vowels” instead of inscribing individually the seven vowels themselves. 6 Kotansky and Spier print {ι} here, but the anonymous reader suggests that instead of deleting ι we consider that this may be a phonetic spelling for the definite article ἡ which is in fact desired for “the seal of Solomon.” 284 NOTES ON MAGICAL TEXTS AND GEMS ————— Greek, Roman, and Byzantine Studies 62 (2022) 280–308 the Seal of Solomon. I have argued elsewhere7 that here, too, the gem cutter (or the handbook he was using) mistakenly confounded two separate and probably contiguous recipes: the first for a gem carved with a lion-headed ouroboros surrounding the invocation that appeared on the gem and that ended with the ÚƵƵƵ symbol and the second for a gem showing a lion- headed Chnoubis and his triple ꞨꞨꞨ symbol, which, as we see on the gem in fig. 5, is sometimes inverted to ƵƵƵ. This mistake arose, no doubt, from the similarity between the two lion-headed snakes and the two triplets, Solomon’s ÚƵƵƵ and Chnoubis’ ꞨꞨꞨ. Here, too, the gem-cutter’s eyes probably jumped from the ÚƵƵƵ in the first recipe to the ꞨꞨꞨ in the second. That a similar error occurred in the engraving of D&D 122 is confirmed by the observation that aside from this gem a corpse never appears below the lion couch in the embalming scenes.8 It may well be that the gem-cutter was supposed to put the image of lion and corpse on the back of the gem, where they would have been encircled (as the mistakenly copied handbook tells us) by χυχ βαχυχ βακαξιχυχ βακαβειχυχ βαδιτοφω βαινχωωωχ. This magical expression, a version of the so-called BAKAKSICHUCH- logos that begins with χυχ βαχυχ, is not, in fact, very common, appearing on only seven other gems, almost all of a similar green medium (serpentine or green jasper):9 SMA 229 (serpentine): χυχ βαχυχ βακαχυχ βαζακαχυχ βακαξιχυχ βαινχωωωχ (alone on the reverse) SMA 230 (green jasper): χυχ βαχυχ βακαξιχυχ βαδη (surrounds figure; reverse blank) 7 C. A. Faraone “The Late-Antique Transfer of Circular Gem-Designs to Papyri and Foil: The Ouroborus and Solomon’s Seal,” in R. Martín Hernández (ed.), The Iconography of Magic: Images of Power and the Power of Images in Ancient and Late-Antique Magic (Leuven 2022) 71–73. 8 The CBd and S. Michel, Die magischen Gemmen: Zu Bildern und Zauberformeln auf geschnittenen Steinen (Berlin 2004) 314, no. 39.6a, both have only eight other examples of the Anubis scene, none of which have the corpse below. 9 See the list in Michel, Die magischen Gemmen: Zu Bildern und Zauberformeln 496. CHRISTOPHER A. FARAONE 285 ————— Greek, Roman, and Byzantine Studies 62 (2022) 280–308 SMA 232 (serpentine): χυχ βαχυχ βακαχυχ βακαξιχυχ βαγητο βαινχωωωχ (alone on the reverse) Michel, Bunte Steine no. 51 (green heliotrope): χυχ βαχυχ βακαχυχ βακαξιχυχ βαζακαχυχ βαδητο φωθφοβηβιβωθ βαινχωωωχ (alone on the reverse) BM 269 (hematite): χυχ βαχυχ βαχυχ (on the reverse) BM 270 (serpentine): χυχ βαχυχ βακαχυχ βακαζαχυχ βακαξιχυχ βαινχωωωχ (alone on the reverse) BM 271 (green stone): χυχ βαχυχ βαχιχχυχ βαζαχυχυχ βακαξιχυχ βαδητοφωθραν χαµωοχευλαµ βαινχωωωχ κατάσχες τοὺς θυµοὺς Τασοι (on the reverse) These seven gems, moreover, have a similar layout (image on the obverse and inscription on the reverse) and a very similar iconography: a solar god with a radiant lion’s head, dressed in a long tunic and raising his right hand flat against his mouth. These gemstones were, moreover, probably all used as amulets against the angry speech of others, a goal indicated by the gesture of the god covering his mouth10 and especially by the additional inscription on the last of the seven gems: “Restrain the angry outbursts of Tasos!” Two other gems can be added to the group. The first once again has the figure of the radiant god on the obverse ( fig. 6) and it has a similar request on the reverse: κατεχέσθω πᾶς θυµὸς πρὸς ἐµὲ Κασισιανόν (“Let all anger against 10 Bonner, SMA 151–152, asserts the following about the gesture: “There is little doubt that this design represents a form of Horus as god of the sun. The rays show the solar character of the deity, the lion head is sometimes given to Horus and the gesture of the right hand is characteristic of Har- pocrates, the young Horus.” S. Michel, Bunte Steine-Dunkle Bilder: “Magische Gemmen” (Munich 2001) 56, agrees, but points out that this interpretation conflates two distinct images of the god: Horus with the lion’s head is the sun at midday, whereas Harpocrates the child sucking on his finger is the sun in the morning. Most telling, however, is that the hand gesture of the solar god on these gems is very different: Harpocrates points his single finger (or some- times his hand) into his mouth, whereas these lion-headed figures hold the hand up parallel to the face in a gesture that aims at covering the mouth. The gesture of Harpocrates was also interpreted as one of silence; for bibliography see R. Mouterde, “Le glaive de Dardanos: Objets et inscriptions magiques de Syrie,” MélBeyrouth 15 (1930/1) 79. 286 NOTES ON MAGICAL TEXTS AND GEMS ————— Greek, Roman, and Byzantine Studies 62 (2022) 280–308 me, Cassianus(?), be restrained!”).11 The second is a large ser- pentine gem in the British Museum ( fig. 7), on the obverse of which we see Hecate (her head obliterated by a later drill hole) holding a flail in her right hand and a sword in her left, while she stands on the back of a prone man (BM 296). These two figures are surrounded by the inscription κατάσχες τοὺς θυµοὺς τοῦ Εὐφράτα καὶ παντὸς ἀντρώπου (“Restrain the angry outbursts of Euphratas and every person”).12 Here we find the same medium (serpentine), the corpse, and the binding of angry outbursts, but the designer has substituted Hecate for the lion. To sum up, then, the inscription on the back of the lost Parisian gem and its green medium suggest that one of the two recipes used to make it was concerned with the suppression of the anger and hostile speech of others.13 Such a suggestion finds support, moreover, in the lower half of the imagery on the obverse of the lost Parisian gem, which belongs to a design well known in another series of gems that show the lion trampling a man and usually placing a paw on his upturned face (see again fig. 3), a scene that is sometimes accom- panied by a female figure who holds her hand to her mouth ( fig. 8) and/or a figure of Helios ( fig. 9), although these images do not seem to be associated with any regular inscription or media: BM 280 (yellow jasper): lion trampling man (obv.); woman hand to mouth and inscription παυσια (rev.) BM 281 (bronze disk; fig. 8): lion trampling man (obv.); woman hand to mouth (rev.) 11 Mouterde, MélBeyrouth 15 (1930/1) 77–80, identifies the figure as the Gnostic goddess Silence (Sigê) and rightly sees that “elle commande le silence,” but the parallels discussed here (most of which he could not have known of) show that this is another example of the lion-headed god. 12 For this reading and full discussion see C. A. Faraone, “Four Missing Persons, a Misunderstood Mummy, and Further Adventures in Greek Magi- cal Texts,” GRBS 61 (2021) 141–143. 13 One can also note that it appears on two curse lead tablets from Car- thage (DT 252.13–21 and 253.22–330) that aim at binding gladiators; in both cases the words are rendered in a column, as they appear on the backs of some of the gems listed above. CHRISTOPHER A. FARAONE 287 ————— Greek, Roman, and Byzantine Studies 62 (2022) 280–308 BM 282 (hematite; fig. 9): lion ridden by radiant Helios tramples man and faces a woman with hand to mouth (obv.) BM 284 (serpentine): lion accompanied by a male god(?) tramples man (paw on back of the head) and faces a figure with hand to mouth (obv.) BM 286 (dark green jasper; fig. 10) lion ridden by radiant Helios(?) tramples man and faces a woman with hand to mouth (obv.) Another gem in this series in Ann Arbor is inscribed with an additional Greek inscription on the reverse “I prevail over you! I hold you!”14 suggesting, as does the trampling action of the lion, that domination is the goal of the whole series. Sometimes the trampling lion appears on the obverse of a gem that has on its reverse a woman with her hand to her mouth in a gesture that is either aimed into her mouth, like the images of Harpocrates, or is flat against her mouth, like that of the solar god discussed above, although she never wears a solar crown ( fig. 8). At other times, however, these two images are combined: the woman stands in this same pose in front of the lion, which tramples the prone skeleton or body and which is on occasion ridden by Helios (BM 282, 284, 286; figs. 9–10). This female figure with hand to mouth is usually thought to represent Nemesis (in which case the hand indicates that she is spitting into her bosom)15 or Sigê (“Silence”; cf. n.10 above). The lower design on the obverse of the Louvre gem, therefore, was probably also used for some kind of suppression, especially since it was placed on a green gemstone in combination with the BAKAKSICHUCH formula. The placement of the female’s hands below the mouth of the lion is, however, the one untraditional part of the second design, one that may suggest she is feeding or comforting the animal. 2. Re-reading some handbook instructions in the Paris Magical Codex A complicated attraction procedure in the Paris Magical Codex (GEMF 57/PGM IV 2943–66) aims at torturing a female 14 SMA 74: κρατῶ σε, ἔχω σε. On another gem of this type (LIM 563) the lion stabs the head with a knife. 15 F. Rausa, “Nemesis,” LIMC VI (1992) 769. 288 NOTES ON MAGICAL TEXTS AND GEMS ————— Greek, Roman, and Byzantine Studies 62 (2022) 280–308 victim with insomnia by making a wax dog and inserting into its head the eyes of a living bat. The recipe then tells us how this dog effigy is to be deposited, using three rare words that have created some confusion: βάλε αὐτὸ εἰς κωθώνιον καινὸν καὶ πιττακίσας αὐτὸ σφράγισον αὐτὸ ἰδίῳ δακτυλίῳ ἔχον⟨τι⟩ κορκοδείλους ἀντικεφάλους αὐτοῖς καὶ κατάθου αὐτὸ εἰς τρίοδον σηµειωσάµενος, ἵνα, ἂν θέλῃς αὐτὸ ἀρεῖν, εὕρῃς. Put it (the effigy) into a new drinking vessel, attach a papyrus strip to it and seal it with your own ring which has crocodiles with the backs of their heads attached and deposit it at a crossroad after you have marked the spot, so that, should you wish to recover it, you can find it. (transl. O’Neil) The general sense of the ritual is clear: make a canine effigy of the female victim, put it in a container, and deposit it at the crossroads. But the devil is in the details. The word κωθώνιον seems to be a rare diminutive of κώθων, “Thracian drinking vessel,” prompting the translation “drinking vessel” in the GMPT, although LSJ say it means “vase” and LMPG, citing this passage alone, translates it as tázon (“bowl”). The second troublesome word, πιττακίσας, is a participle of πιττακίζειν, a hapax which LSJ translate as “attach a label to”; O’Neil quotes LSJ in the note to his similar translation, in which he presumably imagines a wide-mouthed drinking cup with a label attached to its side wall (or perhaps to the wax dog inside), a plausible reading, which is confounded, however, by the next verb in the sentence, σφραγίζειν, which means to “seal” some- thing with a signet ring, originally in the sense of to “enclose” or “seal” the contents of a container, like an amphora, or a document, like a will and testament. But the verb can also mean in a more abstract sense to “authenticate” or “certify” that the contents sealed in the container or document were from a certain person or of a certain quality. It is hard to see, however, how this second, more abstract meaning would make any sense in this rather concrete set of instructions, and we must under- CHRISTOPHER A. FARAONE 289 ————— Greek, Roman, and Byzantine Studies 62 (2022) 280–308 stand the first meaning of the verb: that the ring is used to seal the wax dog inside of the κωθώνιον. The third rare and difficult term is ἀντικεφάλους, which de- scribes the crocodiles that appear on the signet ring used to seal the kôthônion. This adjective is also a hapax, but it is clearly related to the noun ἀντικέφαλον, which, according to LSJ, means “the back of the head” and from which they go on to suggest that the adjective means “back-to-back,” an idea that is picked up in O’Neil’s translation: “with the backs of their heads attached.” The DGE, however, translates it as afrontado (“face to face”) and the LMPG as opuestos por las cabezas (“confronted by their heads”). A similar sealing ritual in this same PGM IV handbook suggests that the word does, in fact, describe some kind of face-to-face confrontation. This recipe is designed to get the skull of a dead man to stop speaking prophecies by “sealing” its mouth (2125– 38):16 Binding seal (katochos sphragis) for those skulls17 that are unsuitable, and also so that they do not speak or do anything at all of these things. Seal (sphragize) the mouth of the skull with the dirt from the doors of (sc. a temple of) Osiris and from mounds (covering) graves. Taking iron from a leg fetter, work it cold and make a ring on which allow a headless lion to be engraved. Let him … trample with his feet a skeleton (the right foot should trample the skull of the skeleton). In the middle of these an owl-eyed cat with its paw on a Gorgon’s head; in a circle around (all of them?) these names: IADÔR INBA NICHAIOPLÊX BRITH. The dirt, in short, will be impressed with the ring and show the same two scenes in reverse, scenes which neatly combine Egyptian and Greek iconography and ritual: 16 For a full discussion, C. A. Faraone, “When Necromancy Goes Under- ground: Skull- and Corpse-Divination in the Paris Magical Papyri (PGM IV 1928–2144)” in P. Struck et al. (eds.), Mantikê: Studies in Ancient Divination (Leiden 2005) 255–286. 17 The Greek has skyphos here and below, which usually means “drinking cup,” but here it must be “skull”; see Faraone, in Mantikê 255–286. 290 NOTES ON MAGICAL TEXTS AND GEMS ————— Greek, Roman, and Byzantine Studies 62 (2022) 280–308 Egyptian Greek Dirt from an Osiris temple from grave mounds Trampling lion with Isis crown (Isis?) owl-eyed cat (Athena?) Trampled skeleton’s skull gorgoneion (Gorgon head) This pair of scenes, the first quite similar to the gem-design with the lion discussed in the previous section, each has its own narrative of triumph of a feline over some kind of head and is designed to re-enforce the goal of sealing the mouth and stopping the skull from speaking. I suggest, therefore, that the image of the two crocodiles in some way also aims at re- enforcing the constraint of the wax effigy in the kôthônion. In this case, there are no obvious parallels among the gems, magical or otherwise, to help resolve the question,18 but a much older Egyp- tian tradition can perhaps be invoked. A number of papyrus amulets from Deir el-Medina dating to the Ramesside period show two or more confronted crocodiles attacking a humanoid or Sethian demon or a pair of them, who are trapped between the opposed snouts of the facing crocodiles. The idea seems to be that the crocodiles, by keeping these figures trapped, will similia similibus prevent the demons from attacking the person who wears the amulet.19 If, as seems the case, the crocodile ring was designed to “seal” the wax dog inside of the κωθώνιον, how was this “sealing” accomplished during the ritual? As it turns out, there are three examples of curses and/or effigies placed within small vases or pots and then sealed. The earliest is a late-classical cooking pot, inscribed with more than fifty names and containing the head and feet of a very small chicken; the contents of the pot were, 18 A. M. Nagy, “Gemmae magicae selectae: Sept notes sur l’interprétation des gemmes magiques de l’époque impériale,” in Atti dell’ incontro di studio 178 n.4, and P. Vitellozzi, “Relations between Magical Texts and Magical Gems: Recent Perspectives,” in S. Kiyanrad et al. (eds.), Methodische Reflexionen zum Spannungsverhältnis zwischen magischem Text und Bild (Heidelberg 2013) 199. 19 J. Dieleman and H.-W. Fischer-Elfert, “A Textual Amulet from Theban Tomb 313,” JARCE 53 (2017) 255–256. CHRISTOPHER A. FARAONE 291 ————— Greek, Roman, and Byzantine Studies 62 (2022) 280–308 moreover, sealed by placing it into a hole excavated in the foundation of a commercial building in the Athenian Agora and then driving a long nail with a very wide head into and through bottom of the pot, so that the nail was fixed to the ground beneath it and so that the head of the nail sealed the pot completely.20 The second example is an inscribed lead tablet from Favianae in Noricum (Mautern an der Donau). The ob- verse reads in Latin:21 “Pluto, or perhaps I should say Jupiter of the Underworld, Aeracura, Juno of the Underworld, summon now quickly the one inscribed below and hand him over to the Manes.” The name of the victim appears upside-down just below this invocation. This tablet had been used to seal the mouth of a small jar (6.5 cm high) made of red clay, which contained some burnt material, and then the whole apparatus was concealed under some stones near an altar that was dedi- cated to the same two deities.22 The final example is a rather late (fifth century CE) ensemble from north of Assiut, Egypt (Suppl. Mag. no. 45): a curse on a papyrus sheet wrapped around two wax figurines, one male and one female, in an embrace, and then inserted into a clay jug (8.0 cm in diameter) that was then sealed with plaster. The precise findspot of this last example is unknown, but the other two were either hidden or buried in the ground. In the end, I suggest that the assemblage from Favianae provides the best comparandum for the PGM recipe under discussion: both use an inscribed papyrus or lead tablet to seal 20 J. Lamont, “The Curious Case of the Cursed Chicken: A New Binding Ritual from the Athenian Agora,” Hesperia 90 (2021) 79–113. The use of birds, lizards, and small animals as effigies in curses is well attested in the Greek and Roman worlds; see C. A. Faraone, “Animal-Effigies in Ancient Curses: The Role of Gender, Age and Natural Behavior in their Selection,” MediterrAnt 22 (2019) 307–334, for a recent discussion. 21 I translate the original reading of R. Egger, “Liebeszauber,” JÖAI 37 (1948) 112–120 (repr. Römische Antike und frühes Christentum II [Klagenfurt 1963] 24–33). 22 Egger (1948) 112 with fig. 25. 292 NOTES ON MAGICAL TEXTS AND GEMS ————— Greek, Roman, and Byzantine Studies 62 (2022) 280–308 the mouth of a jar or vase that contains either an effigy of the victim (the wax dog) or burnt material (Favoniae) that was either the remains of an effigy or more likely some kind of material taken from the victim —e.g., hairs, fingernails, or threads from a garment—and burnt during the ritual. The PGM recipe ends by telling us to “deposit it (the sealed κωθώνιον with the wax dog) at a crossroad after you have marked the spot, so that, should you wish to recover it, you can find it.” The necessity for placing a sign on the spot suggests that in this case as well the whole apparatus was buried or otherwise carefully hidden. How, then, are we to imagine the apparatus described in this recipe? To begin with, we need to alter the translation to reflect the comparanda discussed above: “Throw it (the little dog) into a small vase and after attaching a piece of papyrus (viz., to the mouth of the vase), seal it with an appropriate ring that has (viz., on its bezel) crocodiles facing head to head.” I suggest that the com- mand to “seal it with an appropriate ring” meant that the pittakion was to be molded over the mouth the vase and sealed by pressing the crocodile ring into the plaster, wax, or pitch used to “seal” the papyrus over the mouth of the jug. A very late survival of this kind of cursing and sealing ritual is found in the Sepher ha Razim, a medieval Hebrew handbook of magic that borrows much from late-antique magical recipes in Greek:23 If you wish to give your enemy trouble in sleeping, take the head of a black dog … and a lamella from a strip of (lead) pipe from an aqueduct … Write (the formula) and put (the inscribed lead) in the mouth of the dog’s head and put wax on its mouth and seal it with a ring which has a lion (engraved) on it. Then go and conceal it behind his house or in a place he frequents. If you wish to re- lease him, (take the dog’s head) away from where it is concealed and remove its seal and withdraw the text and throw it into a fire, and he will fall asleep at once. 23 J. G. Gager, Curse Tablets and Binding Spells from the Ancient World (Oxford 1992) no. 114. CHRISTOPHER A. FARAONE 293 ————— Greek, Roman, and Byzantine Studies 62 (2022) 280–308 Here we see the same pattern that we saw in the GEMF 57 recipe under discussion, which is also a spell to induce insomnia: place an effigy of the victim (a wax dog impaled with the eyes of a bat) into a container, seal it with a signet ring, and then conceal it somewhere, but leave open the possibility that the victim can be released by recovering the container, removing the seal, and de- stroying the curse. The GEMF 57 recipe does not spell this out in detail, but it ends with some instructions that suggest the same reversibility of the operation: “deposit it at a crossroad, after you have marked the spot, so that, should you wish to recover it, you can find it.” 3. Re-reading a curse from the hippodrome in Carthage One of the curses found in Carthage’s hippodrome has at the top center, a somewhat schematic drawing of a man with circles drawn over his chest and crossed feet ( fig. 11), which seem to indicate the parts of the body to be bound.24 The image is flanked on each side by three rows of symbols followed by two lines of Greek—the popular magical word SEMESEILAM divided into two sections under the symbols on the left ( fig. 12) and on the right vowels followed by the magical word ABRASAX ( fig. 13). This seems to be a Greek template of sorts, to which two Latin curses have been appended, perhaps at a later time and even by a different hand: the first begins just under SEILAM on the left, running down in a narrow column and then expanding across the tablet under the drawing of the bound man; and the second is at the far right, next to the second set of symbols and it runs down in a much wider column. Both Latin curses begin with similia similibus formulae that are addressed to a singular, superhuman agent, whose name is pre- sumably spelled out by the nearby magical symbols and magical words. We begin with the second curse (the one on the right), because the reading is more secure (spelling regularized): 24 D. R. Jordan, “New Defixiones from Carthage,” in J. H. Humphrey (ed.), The Circus and a Byzantine Cemetery at Carthage (Ann Arbor 1988) I 129–134. 294 NOTES ON MAGICAL TEXTS AND GEMS ————— Greek, Roman, and Byzantine Studies 62 (2022) 280–308 q[uomod]o ped[es] habes l[igat]os, sic et eius Salvi ligate pedes Alumno et Pyro, Polyarce at Lascouuo, obligate pedes OTINOS ut obruant. pre[ndite,] [ob]ligate ped[es] [Ba]tes et Lucife[ro] e[t] – – Just as you keep the feet bound (i.e., here in the drawing on the tablet), so too bind up the feet of him, Salvus, (and those of) Alumnus and Pyrrhus, Polyarches and Lascivius(?). Tie up the feet of each one(?),25 so that they collapse. Seize, tie up the feet of Batê and Lucifer a[nd] – – Here, as Jordan suggested, the first part of the formula (“Just as you keep the feet bound”) probably refers to the effigy depicted in the center of the tablet, whose legs are crossed and encircled to indicate the binding. The other curse (the one on the left) takes a similar form, but is directed at a different part of the victim’s body, once again presumably the place indicated by the circle around the wrists and chest in the upper part of the drawing: [q]uomodo TOMACOS habes ligatos, sic et Concordi OMATE[.] a(l)ligat(e) bracchia, cor, sensum c[um ...]tinor, cum Karcidone, cum Concordio Just as you keep the tomacos bound (i.e., here in the drawing on the tablet), so too, bind up the omate[.] of Concordius, his arms, heart, sense, along wi[th ...]tinor, with Carcido, with Concordius This formula makes the unknown word tomacos the first target of the similia similibus formula, and then adds “arms, heart, sense” as additional points of attack. Since all of the added targets fall within the second circle around the middle of the body, it must be the case that the tomacos resides there as well. Jordan suggested that tomacus is a new word for the lexica, from which was derived the diminutive tomac(u)lum, a word used by Juvenal and other Silver Latin authors to refer to a “sausage.” He ends by asking the question: “does tomaci mean ‘intestines’ here?” This is pos- sible, but such a word must derive, like tomac(u)lum, from Greek τοµή (“a cutting into pieces”) and therefore it refers to the fact 25 It is unclear what otinos means in Latin, but it could be some alternate form of the genitive of Greek ὅστις, here meaning “anyone (i.e., of them)”? CHRISTOPHER A. FARAONE 295 ————— Greek, Roman, and Byzantine Studies 62 (2022) 280–308 that in making sausage, the intestine is divided up and tied into individual sections. The sense here would be some kind of slang, e.g. “may his ‘sausages’ be bound!” I suspect, however, that tomacos is simply a mistake for (s)tomacos, i.e. Latin stomachus, a word borrowed from Greek, which can refer to the “stomach” itself but also to the “esophagus” and to the alimentary canal in general.26 Its appearance here with “arms, heart, sense” is odd, but may perhaps be a collapsed version of the traditional head- to-toe curse, in this case a short list that focuses on the upper body alone. Compare, for example, a second-century CE Greek curse against a stone-cutter from Attica named Paulos (DT 74), which among other things aims at binding his “shoulders, arms, chest, stomach, back” (ὤµους βραχίονα[ς στῆ]θος στόµαχον νῶτ[ον). Indeed, the Carthaginian curse might even be a transla- tion of a Greek original that had, e.g., στόµαχον βραχίονας φρένας καρδίαν, in which case the author took φρένας in the abstract sense of “wits,” rather than the concrete sense of the original “midriff.” In this reading, however, the plural ending of (s)tomacos on the curse from Carthage is a mistake, albeit an error not without parallel in such texts.27 In the second part of the formula, the corresponding word omate[ ] must also represent (s)tomacos and as such is even more corrupt. 26 After this article was submitted, Kent Rigsby informed me that A. Kropp, Defixiones. Ein aktuelles Corpus lateinischer Fluchtafeln (Speyer 2008) no. 11.1.1/37, printed the emendation ⟨s⟩tomac⟨h⟩os without comment and citing only J. Tremel, Magica Agonistica: Fluchtafeln im antiken Sport (Hildesheim 2004) no. 68, who in fact follows Jordan both in printing tomacos without comment and in translating it as “Därme.” Most recently D. Urbanova, Latin Curse Tablets of the Roman Empire (Innsbruck 2018), prints stomachos without brackets and translates it as “entrails” (351, also citing Tremel), but then in her appendix (478–479 no. 140) prints it with brackets (no translation). Clearly someone thought of this emendation earlier, but it is not clear who did so. And as far as I can tell there has never been a defense made of this reading and why it probably does not refer to the entrails or intestines. 27 The anonymous reader asks helpfully whether the plural (s)tomacos might be inspired by Greek φρένας. 296 NOTES ON MAGICAL TEXTS AND GEMS ————— Greek, Roman, and Byzantine Studies 62 (2022) 280–308 4. A unique magical gem in Brisbane In 1983 the R. D. Milns Antiquities Museum at the University of Queensland acquired a magical gem of great interest, whose design is, as far as I can tell, unique ( fig. 14.a).28 Aside from a large chip in the upper right corner, the gem is well preserved. It is a rose-colored stone with mottling in black that has on its obverse a finely carved image of a well-muscled, bare-chested man in an Egyptian kilt, lifting his left foot gracefully in a con- trapposto stance and holding an oversized sistrum in his right hand and in the left an oversized caduceus of Hermes, the staff of which passes beneath his left armpit like that of a riding crop. But out of his neck spring two bearded snakes, one looking toward the sistrum and the other toward the caduceus. At the point where each snake emerges from the neck, we see rods emerging as well, whose tops are bent sharply toward each other and seem to touch. They are perhaps meant to form the letter mu, but, if so, the significance of the letter is unknown. On the reverse ( fig. 14.b) we find vowels arranged in three rows: αη / ααι / αη. The image on the Brisbane gem is unique, but it does follow a pattern seen on a handful of other gems, which depict a bare- chested male figure with a single snake-head, who wears a similar kilt ( fig. 15), but with one exception holds two different attri- butes: a was-scepter in one hand and an ankh-sign in the other: BM 474 (dark brown siltstone; snake is bearded) BM 475 (red jasper; snake is bearded) SMA 262 (ironstone, probably limonite) LIM 271 (greywacke) C. A. Faraone, “Some Magical Gems in London,” GRBS 57 (2017) 413, no. 17 (much worn hematite; spear replaces was-scepter) An outlier in this series is yet another hematite gem, SMA 264 ( fig. 16.a), on which the bare-chested male figure holds similar 28 Many thanks to James Donaldson, the Curator at the Museum, for allowing me to publish the gem (inv. no. 83.010) and discussing various aspects of its design and its acquisition from Christie’s in 1983. CHRISTOPHER A. FARAONE 297 ————— Greek, Roman, and Byzantine Studies 62 (2022) 280–308 attributes, but has two heads, one of a plumed-snake on the left and the other of an ibis with atef-crown on the right. On the reverse ( fig. 16.b) a series of magical names ends with a plea to Chnoubis to help the owner’s digestion. A god with an ibis head alone would be Thoth, of course, but the identity of the snake- headed deity is unclear. Michel lists four more examples of this double type, albeit on different types of stone: green slate, red jasper, and hematite.29 Thus, out of the ten parallel examples two are red jasper and two hematite, an iron ore popular for amulets because of its uncanny ability to “bleed.”30 The other six are stones used less frequently as amulets: ironstone, siltstone, limonite, greywacke, and slate. In this group the unidentified rose-colored stone used for the Brisbane gem is not an outlier. Unfortunately, neither the image nor the vowels on the Brisbane gem give us any clues as to its precise amuletic purpose, although the appearance of similar images on hematite stones and their inscriptions may point to it use as an amulet for stomach prob- lems. 5. A new magical gem in Montreal A green jasper gem from Asia Minor recently added to the antiquities collection of the Montreal Museum of Fine Arts has a standard image of the anguiped on the obverse.31 The rest of its obverse and reverse is filled with magical names that appear as a series of concentric circles. The three rings on the obverse need further comment. They run as follows: Outer ring: Ιαω Αδωναι ταυτολ Αβρασαξ αβρια φριξ αβεραµεωοω Second ring: υθουρεθεεσαννασεθερωνευαλαναθααµ Third ring: βυεγλµσηθσσαµφαουχιουωινωι The three names in the outer ring (Ιαω, Αδωναι, Αβρασαξ) are well known in magical texts and need no further discussion, but 29 Michel, Die magischen Gemmen: Zu Bildern und Zauberformeln 328, Type 46.4. 30 C. A. Faraone, The Transformation of Greek Amulets in Roman Imperial Times (Philadelphia 2018) 94–97. 31 Faraone, Transformation 183, no. 23. 298 NOTES ON MAGICAL TEXTS AND GEMS ————— Greek, Roman, and Byzantine Studies 62 (2022) 280–308 the others are rarer. The words αβρια φριξ seem to be solar names or epithets and show up on inscribed thunderstones that were used as amulets to prevent buildings and persons from being struck by lightning.32 The last word in the outer ring (αβεραµεωοω) is not to be separated from the text in the second ring and the first three letters of the inner ring, which when taken all together preserve a rather corrupt version of the ABERMENTHO-palindrome found on gemstones and in the magical papyri (for better comprehension I have isolated the central nu of the palindrome with dashes):33 Montreal: αβεραµεωοωυθουρεθεεσ-ν-νασεθερωνευαλαναθααµ/βυ LIM 271: αβεραµενθωουλερθθξα-ν-αξεθρελυοωθνεµαρεβα PGM XIV 24: ανεραµενθωουλερθεξα-ν-αξεθρελυωθνεµαρεβα This palindrome seems to have been used in many contexts, for example in invocations of Apollo and other solar deities, but also in an invocation of Seth.34 The gem-cutter who produced the Montreal gem, or the author of the handbook he used, did not seem to realize, however, that the word was a palindrome. The rest of the third ring (εγλµσηθσσαµφαουχιουωινωι) shows little resemblance to the logoi typically found on gemstones, aside from the possible appearance of the name Seth (σηθ) near the start. The two concentric rings on the reverse are inscribed: Outer ring: αβλαναθαναλβα ακρχαµµαχαµαρει ιευωαβ ειµι Second ring: οχωσσµ ιαηιεηιουωηιη σουµαρταυ This sequence begins with two well-known magical words: the αβλαναθαναλβα palindrome (this one correctly rendered) and ακραµµαχαµαρει.35 Both show up together on some gems (e.g. 32 C. A. Faraone, “Inscribed Greek Thunderstones as House- and Body- Amulets in Roman Imperial Times,” Kernos 27 (2014) 251–278. 33 Bonner, SMA 202–203. Only LIM 271 comes close to executing the palindrome perfectly. 34 See Bonner, SMA 202–203. 35 W. M. Brashear “The Greek Magical Papyri: An Introduction and Sur- CHRISTOPHER A. FARAONE 299 ————— Greek, Roman, and Byzantine Studies 62 (2022) 280–308 BM 217) that depict the anguiped, as the Montreal gem does. The following letters ιευωαβ may be a shortened form of a name found on a hematite gem (CbD-72): ιευ-ωεαιωα-ωαβ. The last four letters of the outer ring (ειµι) might signal the use of the “I am” formula that appears sometimes in magical incantations, when the practitioner impersonates a god, e.g. “I am Ieuwab” (cf. SMA 312, “I am Noskamardotenan”). The middle of the second ring is taken up with a series of vowels perhaps to be divided up into smaller groups—e.g. ιαη ιεη ιου ωηιη—followed by the word soumarta, which seems to be the Greek way of ren- dering the Semitic verb “to protect.”36 6. Two gems once in the possession of Toni Raubitschek In the early nineties, my former teacher Toni Raubitschek mentioned in a hand-written letter two gemstones in his pos- session which (he reported) were inscribed respectively απρατος and ηρατος and which he suspected to be of a “magical” nature. I wrote back asking him for more precise descriptions and photo- graphs (or drawings) of the stones or any information about the types of gemstones or the figures depicted on them. Roughly thirty years later I found his letter in a file and I realized that I never received an answer from him. Since he passed away in 1999, I contacted the department and museum at Stanford thinking they had been donated there, but they had no records of such a bequest. The inscriptions are, however, worth further interrogation and inclusion in the ever-growing corpus of magi- cal gems. On magical gemstones single-word inscriptions are usually the names of divinities, demons, or angels, or nonsense inscriptions known as voces magicae. Both of the Raubitschek inscriptions, however, seem to share the same genitive end- ing -ατος which suggests a rarer, but important form of inscrip- tion found on the reverse of magical gems, in which the purpose vey; Annotated Bibliography,” ANRW II.18.5 (1995) 3577–78, and Michel, Die magischen Gemmen: Zu Bildern und Zauberformeln 481–482. 36 Michel, Die magischen Gemmen: Zu Bildern und Zauberformeln 145. 300 NOTES ON MAGICAL TEXTS AND GEMS ————— Greek, Roman, and Byzantine Studies 62 (2022) 280–308 of the stone is indicated by an objective genitive. Bonner, for example, discusses amulets for digestive problems which bear a simple inscription στοµάχου, “for the stomach,” or sciatica amulets inscribed σχιων, a truncated form, apparently, of the genitive ⟨ἰ⟩σχίων, “for the hips.”37 Following this line of thought I suggest that ηρατος on the second of the Raubitschek gemstones may be a mistake for ἥπατος, “for the liver.” There are no direct parallels for the use of this genitive alone on a gemstone, but the genitive occurs twice in more elaborate inscriptions: “Lord Baith Brith … deliver Priscus from the pain of his liver (ἥπατος)!” (LIM 604) and “Ares cut the pain of the liver (ἥπατος)” (BM 385, see fig. 17, where the word appears above and to the right of the god’s head); both of these gems are hematite, on whose obverse the inscription encircles an image of Ares. As for the other inscription, απρατος, I can only diffidently suggest a series of errors for ἕρπητος. Since the word ἕρπης has two meanings, the inscription could indicate either a cure for the skin disease known as “herpes” or “shingles” or a defense against a “snake.” The former is perhaps more likely, given the medical focus of many body amulets, but a possible parallel for the alternative “snake” is a gem (BM 633) inscribed σφεον. Bonner suggested this was a mistake for ὀφέων, arguing in addition that this type of inscription of the name of a harmful animal in the genitive case lies behind the joke at Aristophanes Plutus 885, a passage which refers to the possibility of inscribing a pro- phylactic ring with the phrase “for the bite of an informer” (συκοφάντου δήγµατος).38 7. A new heliotrope with anguiped Richard Levey brought to my attention a magical gem in his possession that preserves a slightly different version of a long formula on a gem published in 1950 by Campbell Bonner (SMA 357). The Levey gem is a heliotrope engraved with the anguiped 37 Bonner, SMA 60 ff. and 72 ff. 38 Bonner, SMA 4–5. CHRISTOPHER A. FARAONE 301 ————— Greek, Roman, and Byzantine Studies 62 (2022) 280–308 on the obverse that displays all the typical features: a rooster head in profile, a bare-chest and snake feet, a flail held in the right hand and a shield in the left. The shield is notable in that its edge faces outwards in profile and this edge is inscribed with a series of vowels ending with the name Iaô: ιλω ιευο/ουω ια/ω (with the common mistake of lambda for alpha in the second letter). The three letters of the solar name φρη (the Egyptian sun god Ra) are scattered around the lower part of his body. A non- sensical magical name was inscribed on the reverse after the oval gem was turned sideways: ζ*ροκοθαρ/α θωβ*ρραβ/υθ. The asterisks indicate a letter that looks like a smaller omicron with a dash or apostrophe coming off the top. There is also an inscrip- tion on the bezel: νεθµυµαωθ µαρµαχθα χαµµαζαχθαθ µρναχαχα. The gem-cutter misjudged the length of the inscription and was forced to overlap the last eight letters. The Bonner gem SMA 357 is a black jasper and has a com- pletely different image on the front: a popular image of Sarapis seated in profile and surrounded by an ouroboros.39 The reverse and bezel of this gem carry a similar inscription in tiny letters that begins on the reverse and goes around the bezel, in this case with less of an overlap. The inscription begins with five of the same magical names that appear on the new gem, but they are in a different sequence and they are missing the penultimate word (underlined) on the new gem: Bonner: νεθµοχλαωµαρακαχθα θωβαρυβαυθ αρναχαχα ζοραοκοθορα Levey: ζ*ροκοθαρ/α θωβ*ρραβ/υθ νεθµυµαωθµαρµαχθα χαµµαζαχθαθ µρναχαχα The individual words can be compared as follows, albeit out of order: 39 See R. Veymiers. “Hileôs tôi phorounti”: Sérapis sur les gemmes et les bijoux antiques (Brussels 2009), for a corpus of the many gems depicting Sarapis. Roughly two thirds of the gems depicting the seated Sarapis carry magical names or words, his types ii.aa 8, ab 3, 12, 13, 17, 27, 32, and 43. 302 NOTES ON MAGICAL TEXTS AND GEMS ————— Greek, Roman, and Byzantine Studies 62 (2022) 280–308 Bonner Levey νεθµοχλαωµαρακαχθα = νεθµυµαωθµαρµαχθα ζοραοκοθορα = ζ*ροκοθαρ/α θωβαρυβαυθ = θωβ*ρραβ/υθ χαµµαζαχθαθ αρναχαχα = µρναχαχα Because of the overlap on both stones, we know in both cases where the inscription begins and ends and we can tell that the Levey gem partially reverses the order of words on the Bonner: Bonner: νεθµοχλαωµαρακαχθα θωβαρυβαυθ αρναχαχα ζοραοκοθορα Levey: ζ*ροκοθαρ/α θωβ*ρραβ/υθ νεθµυµαωθµαρµαχθα χαµµαζαχθαθ µρναχαχα It is curious, finally, that the gem-cutter did not fill up the flat reverse side of the Levey gem first, before continuing on to the bezel, just as we see on the Bonner gem, and that he turned the gem sideways to inscribe the names in ‘landscape’ format; he presumably did so, either because the formulary he was using suggested that the first two words (ζ*ροκοθαρα θωβ*ρραβυθ)40 were the most important names of the anguipede, or because the formulary made it clear that the bezel had to be completely encircled with the inscription. 8. A lost and found gem in the Louvre A rectangular steatite gem (2.1 x 1.7 cm) from Greece ( fig. 18) was purchased by the Louvre in 1889 and has fallen off the radar of most historians of magic since 1933.41 It depicts on its obverse a profile figure without arms below the shoulders (the torso looks like an empty cuirass), wearing a kilt and striding to the right. Its head is leonine, with two facing hook-shaped objects protruding from the top of its head. A line or most likely a scratch seems to run from its nostril upwards on a diagonal. An inscription runs 40 For the second of these words see R. D. Kotansky, “θωβαρραβαυ = ‘The Deposit is Good’,” HThR 87 (1994) 367–369. 41 A. Dain, Inscriptions grecques du Musée du Louvre (Paris 1933) 180–181, no. 206 (Inventory no. Bj 1728bis, Accession no.: MNC 1061). CHRISTOPHER A. FARAONE 303 ————— Greek, Roman, and Byzantine Studies 62 (2022) 280–308 vertically from the bottom left side with three and a half lines in front of the figure and then three more lines on the right side: left side right side ερεορµενφερ ριουρβαεεχερµεν µεργαρµαρβ ουτθλνρεθεξαν ραφορ ωαξτραοευτ Both inscriptions comprise versions of well-known logoi. Those to the left, along with the first five letters on the right, seem to be a version of the SOROOR-formula, with two rhyming segments (they are underlined) added in the middle: Louvre: ερεορ µενφερµεργαρµαρ βραφορ ριουρ Soroor: σοροορ µερφερ γαρ βαρµαφ ριουη ριγξι Some of the letters to the right, on the other hand, preserve a version of another popular magical word, αβεραµενθω, which usually appears as a palindrome:42 Louvre: βαεαερµεν ουτθ λνρθεξαν LIM 271: αβεραµεν θωου λερθεξαν The final line on the right side reads ωαξτραοευτ and does not seem to be a known sequence of letters in a magical name. The logos on the reverse of the Louvre gem is a standard one on gems:43 αυσ λεθρθεµεινω πισσιδαονιβαρ ιµφ The final rho in line 3 is inscribed on the bezel, which also pre- serves the word µαιµαριβα just under the fourth line (ιµφ), suggesting that it might have been part of this sequence. It may be a coincidence that the word αβεραµενθωουλερθεξαν (discussed above) when written out (as it often is) as a palindrome, ends with the letters µαρεβα (see again LIM 271 reverse: αβεραµ/ενθωουλ/ ερθθξαναξ/εθρελυοω/θνεµαρε//βα). There do not seem to be 42 See n.33 above. 43 See e.g. Brashear, ANRW II.18.5 (1995) 3591; Michel, Die magischen Gemmen: Zu Bildern und Zauberformeln 485 and 510. 304 NOTES ON MAGICAL TEXTS AND GEMS ————— Greek, Roman, and Byzantine Studies 62 (2022) 280–308 any other letters on the bezel.44 October, 2022 Department of Classics The University of Chicago cf12@uchicago.edu Figure 1.a-b: D&D 122 Figure 2: D&D 120 Figure 3: BM 280 44 Thanks are due to Roy Kotansky, Attilio Mastrocinque, and Arpad Nagy, who read through earlier drafts, offered many suggestions and saved me from errors. CHRISTOPHER A. FARAONE 305 ————— Greek, Roman, and Byzantine Studies 62 (2022) 280–308 Figure 4: Peiresc drawing Figure 5: BM 305 Figure 6: Mouterde, MélBeyrouth 15 (1930/1) 79 306 NOTES ON MAGICAL TEXTS AND GEMS ————— Greek, Roman, and Byzantine Studies 62 (2022) 280–308 Figure 7: BM 296 Figure 8: BM 281 Figure 9: BM 282 Figure 10: BM 286 CHRISTOPHER A. FARAONE 307 ————— Greek, Roman, and Byzantine Studies 62 (2022) 280–308 Figure 11: Figure 12: Figure 13: upper center to left of image to right of image – Jordan, Carthage no. 3 – Figure 14.a-b: Brisbane gem Figure 15: BM 474 Figure 16.a-b: SMA 264 Figure 17: BM 385 308 NOTES ON MAGICAL TEXTS AND GEMS ————— Greek, Roman, and Byzantine Studies 62 (2022) 280–308 Figure 18.a-b: Louvre gem