Cook ————— Greek, Roman, and Byzantine Studies 62 (2022) 203–238 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/ Philip V and Lysimacheia: An Oath in Gold Brad L. Cook N 202 B.C., as part of his efforts at eastward expansion, Philip V of Macedon formed an alliance with the city of Lysimacheia, brief mention of which is made by Polybius (15.23.7–9, 18.3.11). In 1915, fragments of the treaty, the official public Macedonian copy, were found near Dion and published by Giorgios Oikonomos.1 In 1955, it was Lysimacheia’s turn to reveal physical evidence of this alliance in the form of a stone shield, carved in relief with Philip’s name and the club of Heracles.2 The shield is 33 cm in diameter and, though the top 1 G. Oikonomos, Ἐπιγραφαὶ τῆς Μακεδονίας I—Epigraphai tis Makedonias (Athens 1915), as two separate fragments, A and B, which D. Pandermalis, “Inscriptions from Dion, Addenda et Corrigenda,” in H. Dell et al. (eds.), Ancient Macedonian Studies in Honor of Charles F. Edson (Thessaloniki 1981) 283– 294, realized were contiguous, thus the now standard continous numbering (cf. SEG XXXI 628); Pandermalis also mentioned the discovery in 1975 of an eight-line fragment that he thinks belongs to this inscription, noting the presence of -λεὺς Φιλιππ- and καὶ αὐτονόµων, but the fragment has yet to be published (he speaks, 286 n.17, promisingly of the “Inscriptions of Dion which is under preparation”); in the absence of this fragment, I follow M. Hatzopoulos’ text, Macedonian Institutions under the Kings (Athens 1996) II 21–23 no. 3; reconstructions appear in H. Schmitt, Staatsverträge III 549, and F. Piejko, “The Treaty between Antiochus III and Lysimachia,” Historia 37 (1988) 151–165, esp. 154–155 (SEG XXXVIII 603). 2 Noticed by Louis Robert while working in the Istanbul Archaeological Museum: “Inscriptions des Dardanelles,” Hellenica 10 (1955) 266–282, esp. 266–271 and pl. XXV; it is reported as coming from Bolayır, the site recently identified conclusively as that of ancient Lysimacheia; see M. Sayar, “Lysimacheia. Eine hellenistiche Hauptstadt zwischen zwei Kontinenten und zwei Meeren,” in V. Cojocaru et al. (eds.), Interconnectivity in the Mediterranean and Pontic World during the Hellenistic and Roman Periods (Cluj-Napoca 2014) 363– I 204 PHILIP V AND LYSIMACHEIA: AN OATH IN GOLD ————— Greek, Roman, and Byzantine Studies 62 (2022) 203–238 quarter is broken off, just enough survives of the final sigma of what must have been ΒΑΣΙΛΕΩΣ—a reconstruction made certain by the carved club of Heracles that stretches across the middle of the shield, under which is fully preserved ΦΙΛΙΠΠΟΥ, a layout similar to that found on bronze and gold coinage of Philip V.3 This shield was part of a monument that must have been erected at Lysimacheia to commemorate the establishment of this alliance, the first and only occasion of Philip’s involvement with the city.4 It was, however, a short-lived alliance. In the winter of 199/8, in the midst of renewed war with Rome, Philip’s protecting forces were recalled home, leaving Lysimacheia to be destroyed by marauding Thracians before the year was out.5 In the museum of the University of Mississippi an additional artifact of this short-lived alliance survives, a gold tablet the size of your hand ( fig. 1), inscribed with an epitome of the oath sworn to confirm the alliance between Philip and the Lysimacheians. The goal of this article is publish the gold tablet and to present the comparanda that I have assembled to aid in deciding whether the artifact is of ancient or modern manufacture. 382; cf. A. Lichtenberger, H. H. Nieswandt, and D. Salzmann, “Die hellenistische Residenzstadt Lysimacheia: Feldforschungen in der Zentral- siedlung und der Chora,” in A. Matthaei et al. (eds.), Urbane Strukturen und bürgerliche Identität im Hellenismus (Heidelberg 2015) 163–192, esp. 168 and n.56; R. Boehm, City and Empire in the Age of the Successors: Urbanization and Social Response in the Making of the Hellenistic Kingdoms (Oakland 2018) 52 n.149. 3 S. Kremydi-Sicilianou, “ΜΑΚΕΔΩΝΟΝ ΠΡΟΤΕΣ ΜΕΡΙΔΟΣ: Evidence for a Coinage under the Antigonids,” RN SER. VI 163 (2007) 91–100, esp. 95 and n.20. 4 Among the many uses of shields in sculpted monuments, note the statue base made up of shields found on Delos: A. Reinach, “La base aux trophées de Délos et les monnaies de Philippe Andriskos,” JIAN 15 (1913) 97–142; P. Bruneau et al., Guide de Délos4 (Athens 2005) 201; B. Rabe, Tropaia, τροπή und σκῦλα: Entstehung, Funktion, und Bedeutung des griechischen Tropaions (Rahden 2008) 123–125; J. Camp, “Excavations in the Athenian Agora, 2008–2021,” Hesperia 84 (2015) 467–513, esp. 502 and fig. 31 for a color photo, and the recent discovery in the Athenian Agora, on which see Camp 499–507. 5 Polyb. 18.4.6; ἀνάστατος is the adjective Philip uses. BRAD L. COOK 205 ————— Greek, Roman, and Byzantine Studies 62 (2022) 203–238 Provenance This artifact, now part of the David M. Robinson Memorial Collection of Greek and Roman Antiquities at the University of Mississippi Museum (inv. 77.3.2273), was in Robinson’s collec- tion when he died in Oxford in 1958, where he had lived and taught for a decade following his retirement from his long career at Johns Hopkins University. Half of the collection remained in Oxford with his wife, from whom it was acquired eventually for the University Museum, the other half going to Harvard Uni- versity per Robinson’s will.6 Robinson’s books and papers stayed in Mississippi and are now in the University Archives and the University Museum.7 The earliest known mention of the gold tablet is in the ad hoc inventory of the collection made after Robinson’s death by William Willis, then professor of Classics at the University of Mississippi and the person who had persuaded Robinson to come teach at the University a decade earlier. Repeated searches of Robinson’s papers have uncovered no mention of the gold tablet by Robinson himself. As with the many other artifacts that he purchased through his long life, he usually gives only the vaguest bits of information in his publi- cations. Such is the case with the six of his inscriptions that he did publish,8 which range from “Some years ago there came into 6 For the half of the collection that went to Harvard see G. Hanfmann, The David Moore Robinson Bequest of Classical Art and Antiquities, Fogg Art Museum, Harvard University, May 1 to September 20 (Cambridge [Mass.] 1961), and https:// harvardartmuseums.org/exhibitions/4773/the-david-moore-robinson- bequest-of-classical-art-and-antiquities-a-special-exhibition (accessed 9 June 2022) for an online version of the exhibit. 7 https://libraries.olemiss.edu/cedar-archives/finding_aids/MUM00393. html (accessed 9 June 2022). 8 The six inscriptions: D. M. Robinson, “An Unpublished Inscription from the Collection of David M. Robinson,” in S. Dow, Prytaneis: A Study of the Inscriptions Honoring the Athenian Councillors (Princeton 1937) 158–160; “A New Fragment of the Fifth-Century Athenian Naval Catalogues,” AJA 41 (1937) 292–299; “A New Fragment of an Attic Treasure-Record,” AJP 58 (1937) 206 PHILIP V AND LYSIMACHEIA: AN OATH IN GOLD ————— Greek, Roman, and Byzantine Studies 62 (2022) 203–238 my collection of antiquities in Baltimore through a dealer an inscription said to have been found in Athens” (1937, 158) to “In my collection there is a late Cretan inscription which I secured in 1901 from M. Kalaïdakis…” (1957, 424). In short, as with so many of his artifacts, the sort of data about his acquisition of the gold tablet that we would wish to find seemed to have been retained only in his memory, which died with him. An odd feature about his inscriptions is that of the nearly three dozen that he owned, Robinson published only six.9 Compared to the endless string of publications of the very many vases, sculptures, and, late in his career, jewelry that he purchased, this low rate of publication is very atypical of him.10 This is not the place for a study of the idiosyncrasies of David M. Robinson as an early-to-mid-twentieth century collector of Greek and Roman antiquities, but, since there may be some suspicion that, as a highly experienced epigraphist, he avoided publishing the gold tablet himself because he might have decided that it could be a forgery, there are two known facts. First, he rarely published any of his inscriptions (and note that four of them were published within a year of one another); William Willis, when asked in the 1980s about the atypical rarity of Robinson’s publication of the 38–44, pl. 1; “A Magical Text from Beroea in Macedonia,” in L. W. Jones (ed.), Classical and Mediæval Studies in Honor of Edward Kennard Rand (New York 1938) 245–253; “A New Arcadian Inscription,” CP 38 (1943) 191–199, pl. 1; “A Lost Cretan Decree Found,” in Hommages à Waldemar Deonna (Brussels 1957) 424–427, pl. LX (inscription first published in BCH 17 [1893] 628–629). 9 For a list of nearly all his inscriptions see J. Bodel and S. Tracy, Greek and Latin Inscriptions in the USA: A Checklist (Rome 1997): 46–49 for Harvard, 138– 139 for the University of Mississippi. 10 For a list of his publications from 1904 to 1950 see G. E. Mylonas, Studies Presented to David Moore Robinson (Saint Louis 1951–53) I xxii–xxxiii, though the authorship of Excavations at Olynthus VII The Terra-Cottas of Olynthus Found in 1931 (Baltimore 1933) and of ch. 1 of Excavations at Olynthus XIV Terracottas, Lamps, and Coins Found in 1934 and 1938 (Baltimore 1952) should be credited to Mary Ross Ellingson: see Alan Kaiser, Archaeology, Sexism, and Scandal: The Long-Suppressed Story of One Woman’s Discoveries and the Man Who Stole Credit for Them (New York 2014). BRAD L. COOK 207 ————— Greek, Roman, and Byzantine Studies 62 (2022) 203–238 inscriptions in his collection, wrote: “perhaps because he valued them most highly, he kept them sequestered.”11 Second, an unpublished fact, Willis himself was entrusted by Robinson with publishing the gold tablet, a detail reported to me by Kent Rigsby, Willis’ longtime colleague at Duke University, to whom Willis showed a photograph and transcript of the gold tablet around 1975.12 As far as Robinson’s assessment of the gold tablet is concerned, that he asked Willis to publish it I take as evidence that he judged it to be ancient.13 11 In a letter of March 17, 1987, to Robert Moysey, now retired from the University of Mississippi, but then busy publishing articles on the unpublished inscriptions in the collection: “Three Fragmentary Attic Inscription,” ZPE 78 (1989) 199–207, on 199. Of Robinson’s decades of epigraphical study, consider his dissertation (1904), printed as three articles and as whole with corrigenda, “Ancient Sinope: First Part,” AJP 27 (1906) 125–153; “Ancient Sinope: Second Part,” AJP 27 (1906) 245–279; “Greek and Latin Inscriptions from Sinope and Environs,” AJA 9 (1905) 294–333; and his last, posthumous article, “A New Logos Inscription,” Hesperia 27 (1958) 74–78, pl. 18; between these bookends, among his innumerable publications are many on inscriptions, but note, e.g., on the topic of letter forms, “Macedonica,” in Geras Antoniou Keramopoullou (Athens 1953) 149–158, pl. 13–15, esp. 151–152. 12 Willis, a native of Mississippi who had taught at the University since 1946, only a few years before he persuaded Robinson to move south, had moved to Duke in 1963 after being “forcefully advised” to leave Oxford due to his support of James Meredith and the integration of the University of Mississippi in 1962/3; the quotation is from Ch. Henderson, Jr., “Quorum pars parva fui,” TAPA 131 (2001) 353–362, at 358; see also J. W. Silver, Running Scared: Silver in Mississippi (Jackson 1984), a colleague of Willis’ at the University of Mississippi who says of him not only that he was “the very best teacher” at the University but also that “Willis was unquestionably the faculty leader of committees that often held the chancellor’s feet to the fire” (81); for a summary of his career see G. Davis, Willis, William Hailey, Database of Classical Scholars (n.d.), https://dbcs.rutgers.edu/all-scholars/9241-willis- william-hailey (accessed 9 June 2022). For any speculation as to why Willis never got around to publishing the gold tablet, I suggest that they consider his busy career as a papyrologist, editor of GRBS, and all the other activities described at the DBCS website. 13 On the important topic of modern forgeries of Greek inscriptions, 208 PHILIP V AND LYSIMACHEIA: AN OATH IN GOLD ————— Greek, Roman, and Byzantine Studies 62 (2022) 203–238 Physical characteristics and text The tablet ( fig. 1) is 9 cm wide and 12.5 cm tall, a square with a triangular top, a handheld version of the larger, pedimental bronze plaques that are found throughout the ancient Greek world.14 The thickness of the metal varies from two to three especially on metal, it is often the case that they are exceptionally poor in manufacture or far too flawless; of the latter, consider the golden tiara of Saitapharnes, or Saitaphernes, a stunning forgery purchased by the Louvre for 200,000 francs in 1896; see the exhaustive coverage of the controversy by M. Collignon, “Tiare en or, offerte par la ville d’Olbia an roi Saitapharnès,” Mon Piot 6 (1899) 5–60; for a recent copper forgery of very poor quality, juxtaposed to an authentic, fourth-century B.C. inscription on bronze, see the thorough analysis by W. T. Loomis,“Entella Tablets VI (254–241) B.C. and VII (20th Cent. A.D.?),” HSCP 96 (1994) 127–160, pl. I–IV; and on the gold, whether of poor quality or simply bizarre, the minute (6.4 by 4 cm) rectangle with a “death mask” surrounded by a repoussé inscription referring to King Dropion, the 3rd cent. B.C. Paeonian king, on display in Skopje in the Museum of the Republic of Northern Macedonia, see H. W. Pleket’s assessment in SEG XL (1990) 560 of the original publication in I. Mikulčić and V. Sokolovska, “ИКОНА НА КРАЛОТ ДРОПИОН,” Macedoniae Acta Archaeologica 11 (1987–89 [1990]) 103–110. Of forged Greek inscriptions on stone, or only on paper, and the notorious case of Stavros Mertzidis (1858– ca. 1930), see L Robert, “Hellenica,” RPhil 13 (1939) 97–217, esp. 136–150, and P. Nigdelis, “Απo την ιστορία της έρευνας της αρχαίας Μακεδονίας Συµ- πληρωµατικά στοιχεία για τη ζωή και το έργο των Μ. Δήµιτσα και Σ. Μερτζίδη,” Μακεδονικά 34 (2003/4) 229–249; and for examples farther north, with excellent photographs, see A. Ivantchik, “Some Fake Inscriptions from Olbia and Tyras,” Ancient Civilizations from Scythia to Siberia 21 (2015) 280–301, esp. 297–299. Fundamental reading is A. Chaniotis, “Archival Research, For- mulaic Language, and Ancient Forgeries of Legal Documents,” in A. P. Matthaiou et al. (eds.), ΑΞΩΝ, Studies in Honor of Ronald S. Stroud (Athens 2015) II 669–690; n.b. his view that “in the study of suspicious documents one should disregard the seemingly genuine elements, because they can easily be reconciled with a late fabrication” (675). See also the case studies on forged Latin inscription in the modern era in L. Calvelli (ed.), La falsificazione epigrafica: questioni di metodo e casi di studio (Venice 2019) 49–68, and the related website of the Epigraphic Database Falsae, https://edf.unive.it (accessed 9 June 2022). 14 E.g., the finial-topped Dodona thank-offering, with ΘΕΟΣ ΤΥΧΑ in the BRAD L. COOK 209 ————— Greek, Roman, and Byzantine Studies 62 (2022) 203–238 tenths of a millimeter around the edges, though it is thinner in the lower right section, as seen in an x-ray image ( fig. 2), where the tablet has been stressed and a fracture or tear developed at some time, resulting in a small triangular section breaking off. The tablet weighs 40 grams; XRF scans of the tablet, front and back, give readings of 99.79 to 100% gold. Unlike many of the larger pedimental plaques it has no decorative frame, but the ten lines of Greek text have been centered on the square body of the tablet such that a centimeter or more of undecorated gold surrounds the inscription, which has been scribed, or traced, into the gold with a stylus; a close-up image ( fig. 3) reveals the grooved furrows made by the point of the tracing tool, which was then also used in the manner of a punch to created dot- serifing at the endpoints of every stroke of each letter. Above the text, centered in the pedimental triangle appears in repoussé a profile female head wearing a Corinthian helmet similar to coin images of Athena. Immediately below the last line of the text, also centered, appears a thunderbolt, also in repoussé. Except for a missing crossbar in line 7, which has left an alpha looking like a lambda,15 the twenty-four-word text is clearly pediment (and genitals at bottom), 330s B.C. (Athens, NAM 803, IG IX.12 1750); the naiskos-enshrined Elean decree for Damokrates, with a grape cluster and two axes in the pediment, Olympia, 300–250 (Athens, NAM 6442, I.Olympia 39; Minon, I.dial.éléennes 34); the quartet of proxeny decrees from Corcya, late 3rd cent. (IG IX.12 789–792); the many 3rd cent. plaques from Entella, one of which has a three-plumed helmet stretching from the text into the pediment. On metal inscriptions, see still—as R. Stroud, “A Fragment of an Inscribed Bronze Stele from Athens,” Hesperia 32 (1963) 138– 143, at 142 n.8; L. Robert, Collection Froehner I (Paris 1936) 47–48, “Péri- polarques,” Hellenica 10 (1955) 283–292, esp. 290 nn.1, 2; J. and L. Robert, Bull.épigr. 1961, 154, p.141—the observation by L. Casson, “Early Greek In- scriptions on Metal: Some Notes,” AJA 39 (1935) 510–517, “Much work is to be done yet on metallic epigraphy” (517), which remains true. 15 A common ancient slip, e.g. the ca. 420 B.C. Boeotian skyphos, now in Athens (NAM 803,), with an abecedarium written twice, once with the gamma as Α and also correctly as Λ: L. H. Jeffrey, The Local Scripts of Archaic Greece2 (Oxford 1990) 95 no. 20, pl. 10. 210 PHILIP V AND LYSIMACHEIA: AN OATH IN GOLD ————— Greek, Roman, and Byzantine Studies 62 (2022) 203–238 legible: ὅρκος· σύµµαχος Oath: I will be an ally ἔσοµαι κατὰ τὰ σ- of the Lysimacheians υγκείµενα Λυσιµ- according to the agreed αχεῦσι καὶ ἐάν τι- 4 terms, and if anyone ς ἴηι ἐπὶ Φίλιππον ἐ- goes to war against Philip πὶ πολέµωι ἢ ἐπὶ Λυ- or against the Lysimacheians, σιµαχέ⟨α⟩ς βοηθεῖν the Lysimacheians Φιλίππωι Λυσιµαχ- 8 will aid Philip έας καὶ Λυσιµαχε- and Philip ῦσι Φίλιππον. the Lysimacheians. This brief text presents a version of the oath taken by the two parties to the treaty but one that has been greatly shortened to fit on this hand-sized tablet. This sort of epitomizing can oc- casionally be found in surviving treaties, both epigraphic and literary, in which it is possible to compare the full version of the treaty, listing all the terms of the alliance, and the oath that epitomizes those terms.16 The inscribed copy of the treaty found at Dion, though fragmentary, permits a partial comparison. The two published fragments of the inscription preserve twenty-four lines of text but only the center portion of those lines. In the last seven lines are preserved key words of the oath sworn by the Lysimacheians, with just the beginning of the oath taken by Philip (Hatzopoulos no. 3.17–23): – – τόνδε τὸν ὅρ]κον· vac. ὅρκος Λυσιµαχέων· ὀµνύω Δί[α, Γῆν, Ἥλιον – – – – καὶ τοὺς θεοὺς τοὺς ἐν Σαµοθράι[κηι – – – – ἐµµενῶ ἐν τ]ῇ φιλίαι καὶ συµµαχίαι ἣν πεπόη[µαι – – – – καθότι ἐπισυντέθειµαι ΟΜΟΣΑ– – 20 – – τὴν συµ]µαχίαν τρόπωι οὐθενὶ καὶ ἐὰν Τ– – 16 As A. Giovannini, Les relations entre États dans la Grèce antique (Stuttgart 2007) 233: “Dans la presque totalité des traités conservés, le contenu du serment est un résumé plus ou moins développé de l’accord lui-même, mais il y a deux notables exceptions,” the exceptionally brief oath in Thuc. 5.47.8, discussed below, and the exhaustive oath between Hannibal and Philip V reported by Polybius (7.9.1–17); cf. F. Adcock and D. J. Mosley, Diplomacy in Ancient Greece (London 1975) 221. BRAD L. COOK 211 ————— Greek, Roman, and Byzantine Studies 62 (2022) 203–238 – – κατὰ δύναµιν] τὴν ἐµαυτοῦ· εὐορκοῦντι µ[έν µοι εἴη – – – – Φιλίππου·] vac. ὀµνύω Δ[ία, Γῆν, Ἥλιον – – … the following o]ath: Oath of the Lysimacheians: I swear by Z[eus, Earth, Helios, … … and the gods in Samothrace [… … [I will abide by t]he friendship and the alliance that I have made [… … just as I have subjoined [… … [the al]liance in no way, and if [… … according to] my [ability]; and by keeping my word [may … … Oath of Philip:] I swear by Z[eus, Earth, Helios, … The Dion fragments preserve enough key words of the two oaths, and, though the right and left edges are lost, we can see that the oath of the Lysimacheians took up five and a half lines (17–22); Philip’s oath was likely of similar length and language. The oath on the gold tablet is obviously much shorter, but such epitomizing of oaths is often found in extant inscriptions, a prac- tice that permitted the goldsmith to fit the readily legible essence of the oath on this palm-sized artifact. Epitomizing an oath: examples in earlier and contemporary alliances The alliance made by Athens in 420 B.C. with a trio of Peloponnesian city-states, the Argives, Mantineans, and Eleans, illustrates how an oath may epitomize a long list of specific clauses in an alliance, and this text is preserved in a literary source as well as in a fragmentary inscription. It takes up sixty- eight lines in the OCT of Thucydides.17 A fragment of the Athenian copy of the inscribed alliance also survives, preserving just enough of the right edge of the stone to show that its twenty- five lines correspond to the first forty-three lines of Thucydides’ text.18 Of those lines in Thucydides’ version, the first forty-seven list the many clauses of the alliance, all of which are then encap- 17 Thuc. 5.47.1–12; for detailed discussion and further bibliography see S. Hornblower, A Commentary on Thucydides III (Oxford 2008) 109–120. 18 IG I3 83; Osborne/Rhodes, Greek Historical Inscriptions, 478–404 BC no. 165; on the “few and small discrepancies” between Thucydides’ transcription and the stone see Hornblower, Commentary III 109–112 212 PHILIP V AND LYSIMACHEIA: AN OATH IN GOLD ————— Greek, Roman, and Byzantine Studies 62 (2022) 203–238 sulated in the prescribed oath: “Let the oath be as follows: ‘I will abide by the alliance according to the agreed terms, justly, without causing harm, without deceit, and I will not violate it by any means or device’.”19 With the phrase “according to the agreed terms,” κατὰ τὰ ξυγκείµενα, the oath-takers encapsulated all the many details of the alliance into their epitomizing oath, saving themselves a great deal of time in the process. Similarly, with the prepositional phrase on the gold tablet, “in accordance with the terms made with the Lysimacheians,” this time- and space- saving convention permitted the engraver to express in a small rectangle of gold the weighty duty of aiding an ally. This epitomizing convention continues into the third century and down to the date of our alliance.20 Earlier in the century, for instance, the Athenian inscription of the alliance with Sparta, brought about by Chremonides in the summer of 268, fills nearly seventy lines with the decree authorizing the alliance; the terms of the alliance then take up some twelve lines with another five lines of instructions, after which appears the very succinct oath: “I swear by Zeus, Earth, Sun, Ares, Athena Areia, Poseidon, Demeter, to abide in the alliance that has been made (τεῖ γεγ[ενηµένηι]); for those who keep their oath may there be many good things, and for the oath-breakers, the opposite.”21 The 19 Thuc. 5.47.8, ἐµµενῶ τῇ ξυµµαχίᾳ κατὰ τὰ ξυγκείµενα δικαίως καὶ ἀβλαβῶς καὶ ἀδόλως, καὶ οὐ παραβήσοµαι τέχνῃ οὐδὲ µηχανῇ οὐδεµιᾷ. There are eighteen more lines in Thucydides’ text that delineate who exactly is to swear the oath in each city-state and when, followed by directions for inscrip- tions to be made by the four parties, with a clause about possible amendments appended (5.47.9–12). 20 E.g., IG IX.I2 583 [Staatsverträge III 523] ca. 216 B.C., τὰ συγκείµενα (73) referring to the many details in 23–52, or I.Cret. III III 4 [Staatsverträge IV 605], ca. 205, τὰ συγκείµενα (48) referring to the many details in 5–47; and, later, e.g., Chaniotis, Verträge 59.82–90 [SEG XXVI 1049], 111/0; earlier uses occur as well, e.g., IG II3 412, esp. 5–11, 341 B.C.? 21 IG II3 912.87–90 (269/8), “ὀµ]νύω Δία Γῆν Ἥλιον Ἄρη Ἀθηνὰν Ἀρείαν Ποσειδῶ Δήµητραν· ἐ]µ[µ]ενεῖν ἐν τεῖ συµµαχίαι τεῖ γεγ[ενηµένηι· εὐορκοῦσιν µὲν] πολλ[ὰ κἀ]γαθά, ἐπιορκοῦσι δὲ τἀνα[ντία. On the Chremonidean War see BRAD L. COOK 213 ————— Greek, Roman, and Byzantine Studies 62 (2022) 203–238 prepositional phrase in the text of 420 B.C., “according to the agreed terms,” has been compressed into an attributive parti- ciple, “that has been made.” This usage is found also in a long inscription of about 200 B.C., contemporary with our alliance. Rhodes made an alliance with Hierapytna on Crete; the decree, after more than eighty lines, gives instructions for administering the “customary” (νόµιµος) oath: “I will abide by the alliance and the agreement that has been made (τᾶι γεγενηµέναι) by the people (of Rhodes) with the Hierapytnians without deceit and without excuses; for those who keep their oath may it be well, and for the oath-breakers, the opposite.”22 The phrase “customary oath,” in use for centuries, is usually understood to imply the invocation of deities and such based on the locale,23 but its use here appears to refer to what immediately follows, namely the common use of the “I will abide” (ἐµµένω) clause along with the closing prayer and imprecation. One thing certainly assumed in this epitomized oath is the essential if-then clause of a defensive alliance, which is spelled out earlier in lines 13 to 15: “if anyone makes war on the city or territory of the Rhodians, the Hierapytnians will aid Ch. Habicht, Athens from Alexander to Antony (Cambridge [Mass.] 1997) 142– 149; J. O’Neil, “A Re-examination of the Chremonidean War,” in P. Mc- Kechnie et al. (eds.), Ptolemy II Philadelphus and his World (Leiden 2008) 65–89. 22 I.Cret. III III 3A.89–91 [Staatsverträge III 551; Syll.3 581], ἐµµενεῖν τᾶι συµ⟨µ⟩αχ[ίαι] καὶ τᾶι συντάξει τᾶι γεγενηµέναι τῶι δάµωι ποτὶ Ἱεραπυτνίους ἀδόλως καὶ ἀπροφασίστως· εὐο[ρ]κεῦντι µὲν εὖ εἶµεν, ἐπιορκοῦντι δὲ τὰ ἐναν{αν}τία; cf. transl. Giovannini, Les relations 287–288. 23 The phrase ὁ νόµιµος ὅρκος, or the related ὁ ἐπιχώριος ὅρκος as above in Thuc. 5.47.8, functions as a shorthand for clauses so “customary” that they were instantly filled in by the speakers; see G. Glotz, “Jusjurandum, Grèce,” in Dar.-Sag. 3 (1900) 748–769, esp. 749–750, so too L. and J. Robert “Une inscription grecque de Téos en Ionie. L’union de Téos et de Kyrbissos,” JSav (1976) 153–235, at 222; cf. A. Bayliss, “Oaths and Interstate Relations,” in A. Sommerstein et al. (eds.), Oath and State in Ancient Greece (Berlin 2013) 147– 325, at 163 n.43, focusing on literary texts; and P. Low, Interstate Relations in Classical Greece (Cambridge 2007) 94–95. 214 PHILIP V AND LYSIMACHEIA: AN OATH IN GOLD ————— Greek, Roman, and Byzantine Studies 62 (2022) 203–238 the Rhodians.”24 On the gold tablet this if-then clause is not allowed to be assumed in the “terms agreed” but is spelled out, taking up two-thirds of the text.25 It seems that the show of arms, like the club on the marble shield at Lysimacheia, was essential in every manifestation of this alliance. The simple reciprocity in the action clauses of all the above- cited alliances was the norm, regardless of how the military power of the respective sides varied. In the case of Philip V and Lysimacheia, would the Lysimacheians actually be in a position to give Philip any aid, or is the reciprocity in the if-then clause of the gold tablet merely formulaic? In contrast to the simple reciprocity of the gold tablet’s if-then aid clause, giving Philip and the Lysimacheians equal responsibilities, an earlier alliance made with Lysimacheia by Antiochus I or II (281–261 or 261– 246) presents a more complex scenario. Found at Ilion, the inscription preserves portions of the quite different oaths taken by Antiochus and by the Lysimacheians.26 Antiochus swears, as do the Lysimacheians, to abide by the friendship and alliance that he has made (4 πεπόηµα[ι]), but then he distinctively adds that, “just as I have ag[reed]” (καθότι συντέ[θειµαι]), I [s]hall pre- 24 καὶ εἴ τίς κα ἐπὶ πόλιν ἢ χώραν στρατεύηται τὰν Ῥοδίων ἢ τοὺς νόµους ἢ τὰς ποθόδους ἢ τὰν καθεστα|κυῖαν δαµοκρατίαν καταλύηι, βοαθεῖν Ἱερα- πυτνίους Ῥοδίοις. 25 Also inscribed on the stele with the alliance between Rhodes and Hiera- pytna is an alliance between Hierapytna and Lyttos, I.Cret. III III 3B, ca. 205 B.C. [Staatsverträge IV 603; Chaniotis, Verträge 26], which does not epitomize the oaths and presents them in full, nearly one hundred words, among which appears still the summarizing phrase, “I will abide by the agreed terms” (ἐµµενῶ ἐν τοῖς συνκειµένοις). 26 I.Ilion 45B (now treated as separate from I.Ilion 45A since Sayar, in Interconnectivity 560–561). Lines [1]–22 preserve most of Antiochus’ oath, 23– 37 most of the oath of the Lysimacheians; see Z. Taşlıklıoğlu and P. Frisch, “New Inscriptions from the Troad,” ZPE 17 (1975) 101–106, pl. IV a, b; J.-L. Ferrary and Ph. Gauthier, “Le traité entre le roi Antiochos et Lysimacheia,” JSav (1981) 327–345; S. Burstein, Translated Documents of Greece and Rome III (Cambridge 1985) 29–30; F. Piejko, “The Treaty Between Antiochus III and Lysimachia: ca. 196 B.C. (with a discussion of the Earlier Treaty with Philip V),” Historia 37 (1988) 151–165; J. Ma, Antiochos III and the Cities of Western Asia Minor2 (Oxford 2002) 266–267. BRAD L. COOK 215 ————— Greek, Roman, and Byzantine Studies 62 (2022) 203–238 serve the city [in autonomy and] in democracy [and …] and un- garrisoned [and free from] tribute” (6–10 καθότι συντέ[θειµαι καὶ δι]αφυλάξω τὴν πόλιν [ἐν αὐτονοµίαι καὶ] ἐν δηµοκρατίαι [– –]σ̣αν καὶ ἀφρούρητον [καὶ ἀφορολόγ]ητον), terms appropriate to the predominant power in the alliance. Then follows the if-then clause: “If anyone makes war [either on the city] of the Lysima- cheians or their for[ts or their l]and, I will aid them just as I have a[greed]” (10–13 ἐάν τις πολεµῆι [ἢ τῆι πόλει τ]ῆι Λυσιµαχέων ἢ τοῖς φρου[ρίοις ἢ τῆι χ]ώραι, βοηθήσω καθότι συν[τέθειµαι]). The parallel attack clause posits “[him]” solely as the object of any attack, not uncommon language in the case of monarchies, but a further difference marks each Lysimacheian as promising not to bring aid but that “I will fight as his ally” (31 συµµαχήσω). The noun σύµµαχος, which is in the first line of the gold tablet, had been used in alliances since they first appeared, but the use of the verb form in the Lysimacheians’ version of the oath here could imply something less active than the “I will aid” of An- tigonus; there are, however, alliances in which both verbs are used as a pleonastic pair.27 The if-clause in the Antigonid alliance, “if anyone makes war,” uses the verb of warring, πολεµεῖν; similarly, the alliance between Rhodes and Hierapytna uses στρατεύεσθαι, as does an alliance between Rhodes and Olus, another Cretan city with an important port.28 Since the Dion fragments preserve only ἐὰν τ- of the if-clause of the Lysimacheians’ oath, it is unclear whether it used πολεµῆι, as Piejko proposed.29 The gold tablet uses the long-established phrase “if anyone goes to war,” εἰ ἴηι ἐπὶ τῷ πολέµῳ, which can be paralleled at least as late as 269/8 in an Athenian alliance with Sparta. The oath in that text leaves out the if-clauses completely by speaking succinctly of the “alliance that has been made” (88–89), but the preceding lines gave the 27 The two verbs can be found paired as synonyms, e.g., Staatsverträge III 510 [SEG XXIII 563; Chaniotis, Verträge 13] (ca. 240–221), 2 συ[µ]µαχήσουσι, 3 βοα[θή]σουσι. 28 Staatsverträge III 551 (ca. 201/0), 12–13 στρατεύηται; 552 [SEG XXIII 547] (ca. 201/0), 28 στρατεύηται, 56 [στρατεύηται]. 29 Historia 37 (1988) 153–155. 216 PHILIP V AND LYSIMACHEIA: AN OATH IN GOLD ————— Greek, Roman, and Byzantine Studies 62 (2022) 203–238 full terms of the alliance with paired if-clauses, one after the other:30 And if a[nyone goes to war against the land of the Athenians or [destroys] their laws [or goes to war against the al]lies of the Athen[ians, the Lacedaemonians and their all]ies [will bring aid in full strength, as they are able. And if any]one goes to w[ar against the land of the Lacedaemonians or] des[troys] their laws [or goes to war against the allies of the L]acedaemonians… The verbal clause “goes to war” is restored in 74 but enough survives of 79 to show the continued use of the phrase as on the gold tablet. The pairing of if-clauses, as in this Athenian-Spartan alliance, is what appears most often, but the sort of compressed, single if- clause as on the gold tablet is also found. Two examples are illustrative, both from the northern Aegean. The first, found in the river near Olynthos where it was thrown around 367 B.C., is evidence of an alliance between Grabos of Illyria and the Chal- cidian League, an alliance abandoned before the inscription was even completed in favor of the (ill-fated) alliance made suddenly with Philip II. The fifteen-line inscription is unfinished but unusually so, in that the third, sixth, eighth, twelfth, and four- teenth lines were not inscribed at all when the inscription was abandoned.31 Even with a third of the lines missing, and damage in others, enough of each clause survives to reveal an epito- 30 IG II3 1 912.74–80, ἐὰν δέ τ̣[ις ἴει ἐπὶ πολέµωι ἐπὶ τὴν χώραν τὴν Ἀθην]αίων ἢ τοὺς νόµο[υς καταλύει ἢ ἐπὶ πολέµωι ἴει ἐπὶ τοὺς συµµά]χους τοὺς Ἀθην[αίων, βοηθεῖν Λακεδαιµονίους καὶ τοὺς συµµάχο]υς τοὺς Λ[ακεδαιµονίων παντὶ σθένει κατὰ τὸ δυνατόν· ἐὰν δέ τ]ις ἴει ἐπὶ π[ολέµωι ἐπὶ τὴν χώραν τὴν Λακεδαι- µονίων ἢ τοὺς] νόµους κατ[αλύει ἢ ἐπὶ πολέµωι ἴει ἐπὶ τοὺς συµµάχους τοὺς Λ]ακεδαιµ[ονίων… 31 D. M. Robinson, “Inscriptions from Macedonia, 1938,” TAPA 69 (1938) 43–76, see 44–47 [Staatsverträge II 307; SEG XXXVII 567]; he notes the possibility that the uninscribed lines were painted (45); that the alliance with Grabos was “not consummated” (47) is belied by the inscription, which, though incomplete, would only have been started once the stonemason had a ratified copy of the alliance in hand, see R. M. Errington, A History of Macedonia (Berkeley 1990) 47 n.5. BRAD L. COOK 217 ————— Greek, Roman, and Byzantine Studies 62 (2022) 203–238 mizing of clauses almost as compressed as that in the gold tablet: [θε]ός. τύχη ἀγαθή. [συµµαχ]ίη Χαλκιδεῦσι [καὶ Γράβῳ τῷ βασι]- λεῖ καὶ α̣[ἱ συνθῆκαι. ἐάν] 4 τις ἴηι ἐπὶ Γράβον ἢ ἐπὶ [τὴν χώ]- [ρην τὴν Γράβου] βοηθεῖν Χα⟨λ⟩κιδέας παντὶ σθέ- [νει κατὰ τὸ δυνατόν. ἐὰν δέ τις] 8 ἴηι ἐπὶ Χαλκιδέας ἢ ἐπὶ τὴν χώ- [ραν τὴν Χαλκιδέων — — —]Σ βοηθεῖν Γράβον Χαλκιδεῦσι [παντὶ σθένει κατὰ τὸ δυνα]- 12 τόν. ὅρκους δὲ ὀµόσαι ἀλλή- [λοις. ἀγαθὰ µὲν εὐορκέου]- [σι, κακὰ δὲ ἐπιορ]κ̣έ̣ο̣υ̣σ̣ι̣ν̣. [Go]d. Good fortune. [Allian]ce of the Chalcidians [and Grabos the kin]g and t[reaty]: [if] anyone goes against Grabos or against [the land of Grabos] the Chalcidians will bring aid in full strength [as they are able; and if anyone] goes against the Chalcidians or against the la[nd of the Chalcidians – –] Grabos will bring aid to the Chalcidians [in full strength as he is ab]le; and they will swear oaths to one anoth[er. May there be good to those keeping their oaths and evils to those brea]king them. This sort of compression is taken one step further in a treaty of the late 390s between Amyntas III of Macedon and the Chal- cidians, found in Olynthos in 1844. It begins with a heading that marks it as the Chalcidian copy, “Treaty with Amyntas son of Arrhidaeus,” but then starts again with a heading common to both sides and with clauses that include both sides:32 συνθῆκαι πρὸς Ἀµύνταν τὸν Ἐρριδαίο. συνθῆκαι Ἀµύνται τῶι Ἐρριδαίου καὶ Χαλκιδεῦσι· συµµάχους εἶν ἀλλήλοισι κατὰ πάντας ἀνθρώπου[ς] 4 ἔτεα πεντήκοντα. ἐάν τις ἐπ᾽ Ἀµύν- 32 Hatzopoulos Macedonian Institutions no. 1.1–8 [Syll.3 135; Rhodes/ Osborne, Greek Historical Inscriptions 404–323 BC no. 12]. 218 PHILIP V AND LYSIMACHEIA: AN OATH IN GOLD ————— Greek, Roman, and Byzantine Studies 62 (2022) 203–238 ταν ἴηι ἐς τὴ̣[ν χώρην ἐπὶ π]ολέµοι [ἢ] ἐπὶ Χαλ[κιδέας, βοηθ]εῖν Χαλκιδέ- [ας] Ἀµύ̣[νται καὶ Ἀµύνταν Χαλκιδεῦσιν] 8 Treaty with Amyntas son of Arrhidaeus. Treaty between Amyn- tas son of Arrhidaeus and the Chalcidians. They will be allies with one another with regard to all people for fifty years. If anyone goes against Amyntas, entering into hi[s land in w]ar, [or] against the Chal[cidians], the Chalcidians will aid Amy[ntas and Amyn- tas the Chalcidians]. The compression here gets the message across and saves time and space by not duplicating the common verbiage, which is so very similar to the gold tablet, especially in the if-then clause. Letter forms and dot-serifing Even though the gold inscription consists of a mere twenty- four words, only four letters of the alphabet are absent (Δ, Ζ, Ξ, Ψ). The height of most is 4 to 5 millimeters, omicron and the single theta being only 3 mm in diameter, and a few, such as a solitary beta, stretching to 6 mm. The spacing between letters is fairly equidistant, 1 to 2 mm, though there is oddly a 4 mm gap after the very first character of the inscription, due, I suspect, to a desire to have ὅρκος· σύµµαχος fill that line, serving as a heading of sorts.33 The alphas are straight-bar alphas, with horizontal bars; they all have longer right legs, some only slightly so but five are markedly longer; similarly there are longer right legs on five of the eight lambas, five of the nine mus, one of the three chis, and two of the five kappas. Of the twelve epsilons, all the hori- zontal elements slant down, by ten degrees or so. The one theta is dotted and, like the omicrons, floats in the middle of the line. The ten pis vary the length of their legs: in three, the right leg (hasta) is somewhat shorter; in three, the legs are almost the same length; in four, the left leg is noticeably shorter. The strokes of the fourteen sigmas are consistently angled at about 45º, which are sometimes the same length, though the top and especially 33 There are thirteen to fifteen characters per line, except in the last, which stops nearly 2 cm before the roughly straight right margin. BRAD L. COOK 219 ————— Greek, Roman, and Byzantine Studies 62 (2022) 203–238 bottom strokes are sometimes longer. The three phis have rounded bodies only somewhat smaller than the omicrons. The two omegas have rounded bodies that are slightly larger than the omicrons with feet that are neither horizontal nor angled up as happens in both bronze and stone inscriptions, but angled down.34 How do these letter forms compare to unquestionably genuine inscriptions from around 200 B.C.? Are these consistently straight-line letters, not curved nor lunate, with straight bars on the alphas, with fairly long middle strokes on the epsilons, and with splayed outer strokes on the mus and sigmas, actually features that belong in the fourth century B.C., and earlier? While, on the other hand, if the omicrons and thetas that are smaller than the other letters and the pis with long right hastae are all typical of the Hellenistic period and later, is this as- semblage of letters on the gold tablet some macédoine of forms that no real ancient person could made? Elizabeth Meyer, in her study of a number of the inscriptions from the sanctuary of Zeus at Dodona, most of the third and second centuries and most on metal, advises that we look more closely: she prefaces her analysis of letter forms by warning that “the existing assump- tions—that large, straight letters on stone or bronze date ‘early’, to the fourth century but no later, and that lunate letters (epsilon, sigma, and omega), alpha with broken bar, and pi with hastae of equal length date late, after 232 B.C.—are at best over-simplifica- tions” (emphasis added).35 The many details of her excellent 34 For other downturned feet of omegas, e.g., on stone, see a 3rd cent. B.C. inscription from Cyprus, T. B. Mitford, “Further Contributions to the Epigraphy of Cyprus,” AJA 65 (1961) 93–151, at 133–134, no. 33, fig. 50; and, much later, I.Magnesia 157a, dated 1st cent. A.D.(?). For examples on lead, though much earlier, see the many omegas in the ca. 500 B.C. letter found on Berezan in Dubois, IGDOP 23 (but note that the drawing there has been hori- zontally compressed, see the drawing in Y. G. Vinogradov, “Drevneisheye grecheskoye pismo s ostrova Berezan,” VDI 118 (1971) 74–100, fig. 4. 35 E. A. Meyer, The Inscriptions of Dodona and a New History of Molossia (Stutt- 220 PHILIP V AND LYSIMACHEIA: AN OATH IN GOLD ————— Greek, Roman, and Byzantine Studies 62 (2022) 203–238 work, a case study of relevance by date and geography for in- scriptions of any connection with Philip V, well illustrate how unproductive such oversimplifications are when one fails to examine the many and intertwined features. These features in- clude: (1) changes as well as continuity in lettering over time, (2) variation between public and private inscriptions, (3) variation between media, as well as (4) variation in a single medium due not simply to variation in technique but even to the size of lettering, principles that should be applied throughout the Hel- lenistic period, and throughout the Mediterranean.36 Comparanda, then, from around 200 B.C. and from both Macedonia and Lysimacheia are needed. For the latter, there is, gart 2013) 20. Summaries, such as in Margherita Guarducci, Epigrafia Greca I (Roma 1967) 368–390—compressed to four pages in her synoptic L’epigrafia greca dalle origini al tardo impero (Rome 1987) 81–84—continue to be cited, as in Loomis, HSCP 96 (1994) 139. 36 These principles are especially well presented in B. H. McLean, An Introduction to Greek Epigraphy of the Hellenistic and Roman Periods (Ann Arbor 2002) esp. 40–45, who similarly warns of the use of “allegedly key developments of particular letter forms” (42). On techniques used on stone, furrowing, in which the pointed edge of a chisel is driven at a shallow angle through the stone to form a groove, versus what is termed stem-cutting, in which a chisel of a certain width is hammered, with great care, vertically into the stone to form a groove the width of the chisel, see S. V. Tracy, The Lettering of an Athenian Mason (Princeton 1975) 86–88, and, on (rare) pointillist inscribing on stone, Attic Letter-Cutters of 229 to 86 B.C. (Berkeley 1990) 228; on engraving techniques used on metal, so different in malleability from stone, see J. Kroll, Athenian Bronze Allottment Plates (Cambridge [Mass.] 1972) 24–26; hammering or punching vertically into the metal with chisels of various widths and also with round punches compresses or pushes the metal to the side, which also occurs when driving a chisel with a somewhat rounded edge through the metal to produce a groove, termed ‘tracing’ by professional engravers—and, if the metal is sufficiently malleable, this inscribing can be done with a stylus or even a reed, at least in ‘fresh’ lead (on which see D. Jordan, “A Personal Letter Found in the Athenian Agora,” Hesperia 69 [2000] 91–103, esp. 93); in engraving proper, or incising, metal is actually cut away by driving a very sharp chisel through the metal; lastly, in lieu of a continuous groove, the letters can be formed with a line of dots made by punching with a point, termed pointillist/pointillé or stippling. BRAD L. COOK 221 ————— Greek, Roman, and Byzantine Studies 62 (2022) 203–238 of course, the ΦΙΛΙΠΠΟΥ on the marble shield, though it is carved in relief. The phi has a triangular body that is just under a third of the entire height, a body only slightly smaller than the third phi on the gold tablet, which is oblong, not triangular; the two pis have hastae all of varying length but both right hastae are noticeably longer than the short half-length hasta generally said to be of the time—on stone, that is; the omicron is smaller, just under two thirds the average height of the other letters, also typical of the era, and as on the gold tablet.37 For Macedonian comparanda of inscriptions on stone we have the valuable photographic collection in the Epigraphic Appendix of Hatzopoulos’ Macedonian Institutions. Eight official inscriptions of Philip’s reign bear comparison to the lettering on the gold tablet, especially the alphas, epsilons, thetas, mus, omicrons, pis, sigmas, phis, and omegas, starting with the Dion copy of the alliance.38 In all eight we find straight strokes (not curved) used in the alphas, epsilons, mus, pis, sigmas, and elsewhere, as in the gold tablet; in the Dion copy of the alliance the thetas and omicrons are smaller and set above the baseline, as on the tablet, 37 Photograph in Robert, Hellenica 10 (1955) pl. XXXV, which includes two coins showing very similar lettering, though the body of the phi on the coins is rounded; a newer photograph of the shield appears in Sayar, in Interconnectivity 558. 38 Hatzopoulos no. 3 “Treaty between Philip V and the Lysimachians,” pl. IV (202/1, Dion), cf. D. Pandermalis, Gods and Mortals at Olympus: Ancient Dion, City of Zeus (New York 2016) 92, for a color photograph; no. 9 “Letters of Philip V to Amphipolis,” pl. XII (218, Amphipolis); no. 11 “Letter and boundary settlement of Philip V between Pherai and another Thessalian city,” image in Pandermalis 92 (206–205, Dion); no. 12 “Military diagramma of Philip V from Amphipolis,” pl. XIII–XVII (ca. 200, Amphipolis); no. 13 “Military diagramma of Philip V from Chalkis and Kynos,” pl. XVIII–XIX (221–197, Chalcis); no. 15 “Letter of Andronikos to the sanctuary of the Egyptian deities in Thessalonike and diagramma of Philip V,” pl. XX (187, Thessaloniki); no. 16 “Letter of the magistrates of Amphipolis to the gym- nasium and diagramma of Philip V,” photographs in Ergon 1984, fig. 20–23 (183, Amphipolis); no. 17 “Letter of Philip V to Archippos and hypomnema of the Euiestai,” pl. XXI (181, near the village of Koilas, north of Kozani). 222 PHILIP V AND LYSIMACHEIA: AN OATH IN GOLD ————— Greek, Roman, and Byzantine Studies 62 (2022) 203–238 though in some of the other seven they are only somewhat smaller and in no. 11, of 206 or 205, they are fully as large as the other letters as too are the phis; the right hastae of the pis on the Dion alliance fragments are 70 to 95% the length of the left hasta, on full count averaging 84%, which is longer than the half to two-thirds ratio to be seen on the remaining seven inscrip- tions; unlike the Dion copy of the alliance and the gold tablet, the top stroke of the pis on all the other seven inscriptions, throughout the range of 218 to 181, extend beyond their hastae, to the left and the right; the top and bottom strokes of the sigmas on the Dion alliance are splayed, not as much as on the gold tablet but clearly not distinctly horizontal as on most of the other inscriptions, though on no. 13 they are as splayed as on the Dion alliance; the five phis on the Dion alliance are quite similar to those on the tablet, slightly oblong; the omegas on the Dion alli- ance are, like the omicrons and thetas, smaller and raised above the baseline and the feet are short and horizontal, not angled down like those on the gold tablet. Differences? The middle strokes of the epsilons are consistently shorter than the top and bottom strokes, which is clearly the case in the other seven, especially where the serifs are more pronounced; the hastae of the mus are quite strictly vertical, not slanted as in the tablet. The majority of the letters on the gold tablet, then, including those that typically change over the centuries (alphas, thetas, omicrons, sigmas, phis), are quite similar to what can be seen in these official documents of Philip’s reign. The somewhat odd features on the tablet are: (1) the long middle strokes of the epsilons, generally a mark of an earlier era, as are (2) the splayed hastae of the mus, and (3), typical of a later date, the right hastae of the pis of equal or nearly equal length to the left, and (4) the downturned feet of the omega, a rare idiosyncrasy, as noted above. What, though, does lettering on contemporary metal inscriptions look like? Examples from Dodona, illustrated in the valuable epigraphical appendix to Meyer’s study, are closest geographically. The majority of the bronze inscriptions are made with the pointillé technique but a couple are made by BRAD L. COOK 223 ————— Greek, Roman, and Byzantine Studies 62 (2022) 203–238 punching with a chisel and at least two combine the two techniques, various rounded letters being pointillé while the majority are punched with a chisel. Meyer no. 17, a manu- mission, is such an example, dated by her between 232 and 190, illustrated with a line drawing from Carapanos’ 1878 publi- cation on Dodona, though the drawing does not reveal the pillowing that resulted from the punching.39 In spite of the distortion caused by this pillowing, it appears that most of the straight lines were punched with a chisel that was about 6 mm wide.40 The middle bars of the epsilons are as long as the top and bottom strokes, providing a parallel to the generally long middle bars of the epsilons on the gold tablet, but, one could say, that is an incidental result of the technique. So, too, with the long right hastae of the pis. What, then, of the fact that the Dodona sigmas and mus all have flaring outer strokes, as on the gold tablet? That is a style employed by the scribe of no. 17, as are the pointillé small, elevated omicrons and thetas, which are similar in form though not technique to those on the gold tablet. What of the omegas? The first, in no. 17.2, consists of a hoop made of seven punched dots with one dot on each side as a foot, both of which are perpendicular to the last dot of the hoop, horizontal to the baseline; in line 7, though, an omega does appears to have short, downturned feet similar to those inscribed on the gold tablet, but, in fact, it is difficult to distinguish the right foot on the tablet itself, and the left foot consists of a single dot, 1 millimeter below the baseline; such solitary examples are 39 Meyer, Dodona 152–153 (NAM 467, SGDI 1347); the drawings from Carapanos, though extraordinarily thorough, are not always perfect copies of the actual plaques, as Meyer notes in various cases from her autopsy of nearly all the inscriptions in the appendix, and as I can confirm from working on a few of the Dodona inscriptions now in Athens. 40 Another chisel, nearly 4 mm wide, was used to make the triangular bodies of the phis and the single branch-like limb of the upsilons. 224 PHILIP V AND LYSIMACHEIA: AN OATH IN GOLD ————— Greek, Roman, and Byzantine Studies 62 (2022) 203–238 not yet evidence of a stylistic trend.41 Other rare contemporary comparanda on metal can be found farther afield, such as the four similar bronze proxeny decrees from Corcyra, dated ca. 220 B.C. Images of three of the four are available for comparison, IG IX.1 686 (British Museum) and 685 and 688 (Athens).42 The lettering on these three is markedly like that on the gold tablet, with straight strokes on all those letters that are often said to become curved in the third century, though the three all have broken-bar alphas. There is some curvature, it is true, on the plaque in London, particularly in diagonal outer strokes of the alphas as well as the deltas and lambdas, but the two in Athens maintain distinctly straight strokes in the outer legs of the alphas and in all the strokes of the epsilons, mus, and sigmas.43 41 Of interest as well is a wholly pointillé inscription from Dodona published with a sufficiently clear photo in P. Cabanes, L’Épire de la mort de Pyrrhos à la conquête romaine (Besançon 1976) 558–560, no. 34 and pl. VI, a grant of citi- zenship dated ca. 205 B.C. that does have lunate omegas (ω) but otherwise has ‘square’ epsilons with middle bars frequently of similar length as their top and bottom bar (though an epsilon in line 2 looks lunate, most of those visible in the photograph are clearly square), and splayed four-barred sigmas. The helpful chart in Meyer, Dodona 39–41, lists for Cabanes no. 34 the apparently lunate epsilon in line 2, but it should be noted that the other two epsilons in the same line, and the other five visible epsilons, are all square; the chart accurately shows a square sigma and a lunate omega, which 19 n.30 correctly noted, “lunate epsilon and omega only,” but 21 n.34 has accidently printed “epsilon and sigma.” 42 For 686, https://www.britishmuseum.org/collection/object/G_1824- 0499-18 (accessed 8 June 2022), and G. Manganaro, “Metoikismos-metaphora di poleis in Sicilia: il caso dei Geloi di Phintias e la relativa documentazione epigrafica,” ASNP SER. III 20 (1990) 391–408, tav. LXXVI–LXXIV; for 685 and 688, Manganaro tav. LXXIX. 43 The observation that the Corcyra plaques as a set are marked by “‘straight’ strokes more curved, extension of horizontal strokes of gamma, epsilon and pi beyond their verticals or vice versa,” Loomis, HSCP 96 (1994) 139, is not true of the two in Athens, which illustrates all the more Loomis’ observation that there is “considerable variation in shapes even within this Corcyraean corpus” (139). BRAD L. COOK 225 ————— Greek, Roman, and Byzantine Studies 62 (2022) 203–238 But, though the engraving technique on these inscriptions on metal appears generally to be by punching with a chisel, what would be ideal is to have comparanda on more malleable metals on which the inscribing could be done with a stylus, with the result that the lettering may be affected by the technique. In the absence of any comparanda on gold for a regional, contem- porary metal inscription made by this simple technique of inscribing, consider, lastly, a lead plaque from Dodona, dated between 219 and 167 or 190, illustrated by Éric Lhôte’s careful drawings based on microscopic analysis.44 The lettering varies in size from 1 to 4 mm, with some omicrons as small as half a millimeter, but throughout the nearly calligraphic lettering, as Lhôte is tempted to call it, we see straight strokes in the alphas, epsilons, mus, sigmas, with outer strokes of the latter two splayed, long middle bars on the epsilons, pis with right hastae that are often nearly as long as their left, all looking very similar to the lettering on the gold tablet, though, it is true, the feet on the omegas do not turn down in the slightest. Mistakes, both in ancient inscriptions and in forgeries, are always interesting. The one visible mistake in the lettering has been mentioned above, the second alpha in line 7 without a crossbar, an absence that is relatively common in ancient texts, which one could interpret as the mark of a terribly clever forger or of a typical ancient inscriber. An X-ray of the tablet, however, reveals an earlier mistake in the second letter in line 3 ( fig. 2): the one gamma in the inscription originally was given a right leg, making it a pi; this error was corrected, the careful smoothing of the gold making the slip nearly invisible. With a few of the diago- 44 Originally published by A. I. Antoniou, Δωδώνη. Συµβολή Ηπειρωτών στην ανοικοδόµηση κτισµάτων του Ιερού της Δωδώνης (Athens 1991), but see the edition by É. Lhôte, “Nouveau déchiffrement d’une petite plaque de plomb trouvée à Dodone,” in P. Cabanes et al. (eds.), L’Illyrie méridionale et l’Epire dans l’Antiquité IV (Paris 2004) 113–131, esp. his microscopically precise drawings, 114–115, and his detailed description of the technique and forms of the letters, 119–121. 226 PHILIP V AND LYSIMACHEIA: AN OATH IN GOLD ————— Greek, Roman, and Byzantine Studies 62 (2022) 203–238 nal strokes, as well, the inscriber appears to have slipped a bit and overshot the intended length. This is evident in the alpha that starts line 4 ( fig. 3): its right leg, instead of being about 4 mm like the left leg, is over 6 mm and the intended length is marked by the terminal dot serif, beyond which the stroke extends for another millimeter.45 This dot-serifing, as I have termed it,46 which becomes ubiqui- tous on coins in the Hellenistic period, can be found already in the sixth century B.C. on the epsilon of Athens’ ΑΘΕ. Later in Athens, they appear in the fourth century as the terminal “punched dots” found on certain Athenian bronze pinakia (Kroll’s Class VI), which are somewhat later in the fourth century B.C. than the well-known “pierced hole” serifs on the pinakia (Classs V) dated to the second quarter of the fourth cen- tury.47 If the Athenians first developed pierced-hole serifing to obviate forgery, as seems likely,48 it soon became a decorative 45 E.g., line 9, fourth character, a kappa has a right lower leg that is 5 mm long with the dot serif leaving 2 mm as a tail; line 6, the omega in the middle of the line has an overshot left leg. 46 There is, it seems, no standard term for this feature in epigraphy, numismatics, or the study of ancient gems: on gems, Gisela Richter speaks of “the little balls at the ends of the strokes of the letters,” Catalogue of the Engraved Gems, Greek, Etruscan, and Roman (Rome 1956) xxxiii; on coins, Percy Gardner calls them “the round dots in which letters of inscriptions often terminate,” The Types of Greek Coins (Cambridge 1883) 21; and J. Hartmann, “knob-like serifs,” Greek Numismatic Epigraphy (Chicago 1969) 16; on the Hellenistic marble copy of the Themistocles Decree from Troizen, Sterling Dow describes “most of the serifs” as “knob-like,” “The Purported Decree of Themistokles: Stele and Inscription,” AJA 66 (1962) 353–368, at 356; of inscriptions on bronze, Ettore Ruggiero calls them “puntolini,” Catalogo del Museo Kircheriano (Rome 1878) 60; in discussing the bronze pinakia, Kroll describes the lettering on the bronze “candlestick” from Dodona as “serifed with punched dots,” Athenian Bronze Allotment Plates 27 n.21; and Loomis, HSCP 96 (1994) 139, says of the letters on Entella C1 that the “ends are marked by slightly enlarged dots.” 47 Kroll, Athenian Bronze Allotment Plates esp. 24–31 and 31–33, on those pinakia with “partially punched holes.” 48 Kroll’s Class V, Athenian Bronze Allotment Plates 26–31 and nos. 78–113. BRAD L. COOK 227 ————— Greek, Roman, and Byzantine Studies 62 (2022) 203–238 feature, appearing on the ends of strokes within letters, in the middle of crossbars, here and there on round letters.49 The fourth-century pinakion from Sinope, with its distinct lettering and dot-serifing, marks a growing use of this artful finishing of letters,50 a feature added to embellish a variety of metal inscrip- tions across the Mediterranean, both through the fourth and third centuries to the time of the gold tablet and beyond.51 This parallels the increasing use of serifed lettering on stone inscrip- tions, but the use of dot-serifing on gold makes the reflective text stand out even more. A morphological peculiarity? The language and especially the lettering of the tablet are just what one would expect for an artifact of ca. 200 B.C. A question could be raised, though, about the accusative plural ending -έας with the ethnonym Λυσιµαχεύς, which appears twice, in lines 7 and 9, the second of which has ‘typographical’ lambda for alpha. The -έας ending is, of course, the accusative form that we all learn as the standard Classical form, but by the end of the fourth century B.C. the alternate form -εῖς had appeared, and slowly 49 See the charts in Kroll, Athenian Bronze Allotment Plates 28; note also those with alternating pierced and non-pierced dot-serifing, nos. 80, 84, 91, 94. 50 Now in Paris, Kroll, Athenian Bronze Allotment Plates 270–272; cf. his nos. 17, 62, 83 (with earlier piercings), 140. 51 E.g., on gold, the Φιλίστη gold leaf found at Aigai/Vergina, listed as “Hellenistic period,” A. Bernabé and A. Jiménez San Cristóbal, Instructions for the Netherworld: The Orphic Gold Tablets (Leiden 2008) 164, 267–269; on the im- mense bronze tables from Heracleia in south Italy IG XIV 645, usually dated late 4th/early 3rd cent. (F. Coarelli dates the tables on historically arguments ca. 350, “Problemi e ipotesi sulle Tavole greche di Eraclea,” in E. Greco (ed.), Siritide e Metapontino: Storie di due territori coloniali [Naples 1998] 281–290); bronze plaque for statue base of Philip of Arcadia I.Olympia 174, early 3rd cent.; Entella plaque C1 (VI Nenci) IGDS 209, 254–241? B.C.; Arcadian proxeny decree NAM 14613/2 (not published), 250–200, Athens; Γλαύκων dedication of a “candlestick” at Dodona LGPN IIIA 99, 3rd/2nd; Sicilian(?) decree honoring Pompey IGUR I 4, 1st cent. B.C.; and a gold ring engraved with salvia Boston, MFA 65.620, 1st/2nd cent. A.D. 228 PHILIP V AND LYSIMACHEIA: AN OATH IN GOLD ————— Greek, Roman, and Byzantine Studies 62 (2022) 203–238 pushed -έας nearly out of existence, though later Atticist authors greatly revived its use. In inscriptions, at least in Attica, it has been stated that “occurrences of -έας in the third century are very rare” and “after 200 B.C. only -εῖς is attested in prose” inscriptions.52 Morphological developments like this are not unique to inscriptions and can be found in literary texts.53 Consider a string of accusative plural ethnonyms in Polybius (no textual variants are recorded in published editions or com- mentaries): Κορινθίους, Φωκέας, Λοκρούς, Εὐβοεῖς, Ἀχαιοὺς τοὺς Φθιώτας, Μάγνητας, Θετταλούς, Περραιβούς (18.46.5).54 The Phocians are given the older -έας ending but the Euboeans have the ‘newer’ -εῖς.55 This variation in the surviving manuscripts of Polybius, manuscripts produced centuries later, illustrates the actual variation in morphology in the second century B.C. itself, a situation paralleled in a very different, and peculiar, category of texts, documentary papyri from Egypt,56 but best proven by 52 L. Threatte, The Grammar of Attic Inscriptions II (Berlin 1996) 246. 53 Though on Atticizing tendencies see G. Horrocks, Greek: A History of the Language and its Speakers2 (Chichester 2010) esp. 133–138; L. Kim, “The Literary Heritage as Language: Atticism and the Second Sophistic,” in E. Bakker (ed.), A Companion to the Ancient Greek Language (Chichester 2010); J. Kazasis, “Atticism,” in A.-F. Christidis (ed.), A History of Ancient Greek, from the Beginnings to Late Antiquity (Cambridge 2007) 1200–1217; A. Dihle, Greek and Latin Literature of the Roman Empire (London 1994) esp. 53–59. 54 Searches were run for these ethnonyms: Ἀλαβανδεύς, Ἀλεξανδρεύς, Ἁλεύς, Ἀντιοχεύς, Ἀστυπαλεύς, Δωριεύς, Ἐρετριεύς, Εὐβο(ι)εύς, Θεσπιεύς, Λα(ο)δικεύς, Κορωνεύς, Λεβαδεύς, Μαντινεύς, Μεγαρεύς, Μελιτεύς, Μηλιεύς, Μυλασεύς, Πεδιεύς, Πεδνηλισσεύς, Πριηνεύς, Σολεύς, Φωκεύς, Χαιρωνεύς, Χαλκιδεύς. 55 By comparison, in Pausanias, an Atticizing author of the second century A.D., the Phocians appear in the accusative plural ten times, nine times as Φωκέας, and once as Φωκεῖς (9.40.12); there are no accusative plural forms of Euboean in Pausanias; but compare, e.g., the appearance in our modern text of Δωριέας in eight passages, but of Δωριεῖς in one (10.8.2). 56 See F. Gignac, A Grammar of the Greek Papyri of the Roman and Byzantine Periods II (Milan 1981) 86–87; and, though ethnonyms in -ευς are not com- mon in the documentary papyri (except, of course, Ἀλεξανδρεύς), variation in BRAD L. COOK 229 ————— Greek, Roman, and Byzantine Studies 62 (2022) 203–238 inscriptional evidence beyond the borders of Attica. We need only cross from Attic to Euboea to find an inscribed example of -έας from the third century B.C. An inscription from Eretria about the Dionysiac technitai uses the accusative plural Εὐβοέας.57 Well into the second century B.C., Εὐβοιέας appears on another Euboean inscription, this one from Chalcis honoring Ergotimos son of Aristotle.58 Of other forms of “Euboeans” only rare examples are extant of the nominative plural Εὐβοιεῖς or Εὐβοεῖς, many uses of the genitive plural, two of the dative plural, but not a single example of the ‘new’ accusative plural -εῖς is found in extant inscriptions. Likewise with the ethnonym “Megarian,” the (slight) extant evidence reveals the continued use of the traditional -έας ending in Μεγαρέας in an inscription (tentatively) dated to the second century B.C.,59 and not a single example of the accusative plural -εῖς can be found in surviving inscriptions. Across the Aegean, still in the second century, the traditional ending in -έας is seen on two inscriptions, one naming the people of Mylasa (Μυλασέας)60 and another the people of the two accusative plural endings can be seen over the centuries in such nouns as γραµµατεύς and its compounds, ἱερεύς, ἱππεύς, etc. 57 IG XII.9 207.17 (280–240 B.C.), K. Kourouniotes, “Ἐρετρικαὶ ἐπιγρα- φαί,” ArchEph (1911) 1–38, pinax 1 (though the photograph is not absolutely clear, not one of the many studies of this inscription has questioned the reading); for the dating given here, in place of the traditional date of 294– 288, see Kent Rigsby, “On the Early Technitai of Dionysus,” Studi ellenistici 31 (2017) 283–286. 58 IG XII.9 898.3; for a drawing, with its “lettres ornés de petits apices,” see A. Joubin and A. Wilhelm, “Inscriptions de Chalcis,” BCH 16 (1892) 90–120, at 100. 59 IG VII 19.11: “2nd century B.C.?” Ph. Smith, The Archaeology and Epigraphy of Hellenistic and Roman Megaris, Greece (Oxford 2009) 180–181, no. 19. 60 I.Mylasa 643.4; on the dating, generally given as 2nd cent. B.C., see N. Unwin, Caria and Crete in Antiquity: Cultural Interaction between Anatolia and the Aegean (Cambridge 2017), esp. 141 for earlier bibliography; for a drawing, clearly showing -ΕΑΣ, see LBW no. 382; Μυλασέας is also restored in I.Mylasa 632.6, 652.8, and 657.6. The acc.pl. in -εῖς is more common, e.g., I.Labraunda 230 PHILIP V AND LYSIMACHEIA: AN OATH IN GOLD ————— Greek, Roman, and Byzantine Studies 62 (2022) 203–238 Priene (Πριηνέας).61 In the case of these two ethnonyms the use of the accusative plural -εῖς does appear more often than -έας, but these two examples, both of the second century, prove that continued use of the traditional -έας around the Aegean. Perhaps one could suggest that the author of the text of the gold tablet was morphologically traditional or conservative in using -έας ending; perhaps, though, it signals an archaizing leaning on the part of the goldsmith. Parasema–emblems–insignia The helmeted head and thunderbolt on the tablet look to be civic insignia, emblems, or, to use the ancient term, parasema (sing. parasemon).62 Civic inscriptions, along with an array of other ancient artifacts, are often marked with a symbol related to the persons honored or otherwise involved in the text. These para- sema are best known from coins—the above-mentioned dikast’s ballot, reported as found in Sinope, is, in fact, stamped at right with the same image that appears on Sinope’s classical coinage, a sea eagle bearing off a dolphin63—but they appear on a variety of objects, including inscriptions, some of which present note- worthy parallels to the gold tablet. Many of the pedimental metal plaques showcase parasema in their pediments. A fragment of a fourth-century Aetolian proxeny decree for an Athenian is topped by a tiny owl three centimeters tall, which presumably 1.3, 4 (240 B.C.); I.Delphinion 146.43 (209/8); I.Magnesia 93a.5, 20 (2nd cent. B.C.); I.Mylasa 633.6–7 (end 2nd/early 1st); I.Priene B-M 75.29–30 (early 1st cent. B.C.). 61 I.Priene B-M 122.21 (ca. 135 B.C.), with a clear photograph. The form Πριηνεῖς can be found in other 2nd-cent. inscriptions, e.g., 8.9 (ca. 200), 110.23–24 (190s), 121.31 (200–150); I.Magnesia 93a.26 (2nd cent.). 62 On the term see S. Killen, Parasema. Offizielle Symbole griechische Poleis und Bundesstaaten (Wiesbaden 2017) 1–2, an invaluable study, with exceptionally good images, which does not, however, include parasema of kings, though it does include poleis within their realms, such as Lysimacheia. 63 See Killen, Parasema 213, Cat. Sin i1, Taf. 18.10; Kroll. Athenian Bronze Allotment Plates 270–272. BRAD L. COOK 231 ————— Greek, Roman, and Byzantine Studies 62 (2022) 203–238 stands for the honorand’s polis.64 A much larger owl is cast in relief on a fourth-century proxeny decree from Corcyra, now in the British Museum, honoring Dionysios, son of Phrynikos, of Athens.65 Instruments and tools are likewise found as parasema on these bronze plaques, such as the kithara that tops a bronze plaque from Lousoi honoring Olympichos of Charadra (near Delphi) as a proxenos and as a thearodokos, a person who provides hospitality to festival heralds, the kithara signaling possibly his role as a thearodokos rather than a topographical relationship.66 A pair of double-headed axes artfully flanking a grape cluster, parasema found on the coinage of Tenedos, grace the center of the pediment on a bronze Eleian decree honoring Damokrates of Tenedos, dated to the first half of the third century B.C.67 A caduceus appears at the bottom of a small bronze Aetolian 64 Athens, NAM X14613/1 + X14613/2, IG IX.12 91.3 (left frag.) + E. Mastrokostas, “Inschriften aus Ätolien, Akarnanien und Westlokris,” AthMitt 80 (1965) 152–159, pls 60–64, esp. 152 (right frag.); see Killen, Parasema 175, Cat. Ath IIk1, Taf. 7.2. 65 London, BM inv. 1868,0110.3, IG IX.I2 786, Tab. V; see Killen Parasema 175, Cat. Ath IIK2, Taf. 7.4. A third owl appears on a fragment of a bronze inscription from Olympia, I.Olympia 819 (3rd cent. based on letter forms), but only just enough letters of the first line preserve the titular invocation, “God. Good Fortune,” and, probably, “people,” but whether of Athens or another city is not preserved; Killen Parasema 175, Cat. Ath IIIK1, Taf. 7.6. 66 Athens, NAM 15403, IG V.2 389 (4th/3rd cent.), see T. Ritti, Sigle ed em- blemi sui decreti onorari greci (Rome 1969) 294–295, no. 43; J. Ma, “A Horse from Teos,” in P. Wilson (ed.), Greek Theatre and Festivals (Oxford 2007) 241 n.46, sees the image connecting to a musical role; both the kithara and lettering are in repoussée. Also from Lousoi is a bronze plaque with a grazing fawn in the pediment, first half of the 3rd cent. B.C., Athens, NAM 15400, IG V.2 392; it is a proxeny decree for someone from Pharai, but whether Pharai in Achaea, Messenia, or Laconia is unclear, as is the connection to a deer; Ritti 294, no. 42. 67 I.Olympia 39; see Killen Parasema 220, Cat. Tene k1, Taf. 21.11; on the coinage of Tenedos the ax is prominent and the grape cluster much smaller, and sometimes absent. 232 PHILIP V AND LYSIMACHEIA: AN OATH IN GOLD ————— Greek, Roman, and Byzantine Studies 62 (2022) 203–238 plaque honoring as a proxenos Laalkos of Pheneos, an Arcadian city that put Hermes’ caduceus on its coinage.68 A bronze plaque from Theisoa in Arcadia (early 2nd cent.), a proxeny decree for Thymon of Thelpousa, has a parasema at the end of the text, a thunderbolt, made of a great many dots, as is the lettering in the plaque, and flanking the thunderbolt are ΘΕΛ at left and ΦΟΥ at right; though the few known coins of Thelpousa have the head of Demeter or Helios, not Zeus nor his thunderbolt, the engraver surrounds the image with the name of the honorand’s city.69 This sampling of parasema, all on metal inscriptions, presents variation in technical and artistic styles, as do the inscriptions themselves. And though the logic in choosing the particular parasema relative to each inscription seems to vary, their usual role is to signal a connection to deities and their power. Just as the watchful owl of Athena and the thunderbolt of Zeus serve as civic as well as universal markers, so the head of Athena and thunderbolt on the gold tablet can be read as both. Athena’s helmeted head is best known on coinage of Athens and Corinth since the sixth century, in an Athenian or Corinthian helmet respectively. But consider Alexander’s use of Corinth’s helmeted Athena, for instance, borrowed, perhaps, “to remind the Greeks that Alexander was leader of the Corinthian league.”70 And the whole of Athena, seated or striding, appears on some coins of his Successors. Lysimachus, the founder of Lysimacheia, minted a 68 Athens, NAM 13671, IG IX.12 22 (mid 3rd cent. B.C.); see Killen Parasema 169, Cat. Phe Ik1, Taf. 4.10; on Hermes at Pheneos, see Paus. 8.14.10. 69 Athens, NAM 14767/1, IG V.2 511; Hiller, Syll.3 623A, suggests on the thunderbolt, “Urbis signum, quia fulgus θάλπει”; Ritti, Sigle ed emblemi 299, sees it as referring generally to Zeus, noting the importance of Zeus Lykaios in Arcadia. Another weapon, a spear, or spearhead, appears at the top of a very fragmentary bronze decree from Arcadia, proxeny for four Aetolians (second half 3rd cent.), Athens, NAM 14613/2, BCH 38 (1914) 454–457, no. 2, fig. 3 [SEG XXXIII 317], Killen, Parasema 159–160, Cat. Ait Ik1, Taf. 2.4; a spearhead commonly appears on coinage of the Aetolian League. 70 K. Sheedy, Alexander and the Hellenistic Kingdoms: Coins, Image and the Creation of Identity (Sydney 2007) 44. BRAD L. COOK 233 ————— Greek, Roman, and Byzantine Studies 62 (2022) 203–238 seated Athena, after 297 B.C., an image used far into the third century, and Antigonas Gonatas, grandfather of Philip V, regularly put a striding Athena, wielding a thunderbolt, on his silver coinage and put Athena’s head in a Corinthian helmet on some of his bronze coinage. The ubiquity of Athena’s head wearing a Corinthian helmet grows in the Hellenistic period, appearing even on Athens’ bronze coinage in certain decades.71 Lysimacheia, which minted its own bronze coins, used a lion, its parasema,72 on the reverse, and on the obverses can be found the head of Lysimachus, Heracles, Demeter, the city itself (wearing a turreted crown), and, quite often, Athena in a Corinthian hel- met.73 On the other hand, Philip V, throughout his long reign, used the heads of Zeus, Poseidon, Apollo, Artemis, Helios, Perseus, and of himself, but not the head of Athena.74 The thunderbolt was employed as a symbol by many, but most famously and for centuries, on the coinage of Elis. Epirote and Macedonian rulers, whether because of the shrine at Dodona or purported descent from Zeus, were partial to the thunderbolt, and Philip V was no exception, as his many decades of coins show. The thunderbolt at the bottom of the gold tablet is notably centered on the word immediately above it, ΦΙΛΙΠΠΟΝ. What is even more notable is that the thunderbolt is not parallel to the text: it is canted up to the right by 22º, looking as if it is pushing its right upper ‘tendril’ into the two pis of Φίλιππον. In addition, of the four vertical strokes of those two pis, the right hasta of the 71 O. Mørkholm, Early Hellenistic Coinage (Cambridge 1991) 87, and esp. J. Kroll, The Athenian Agora XXVI The Greek Coins (Princeton 1993), e.g., 33–34, 44–47, pls. 5–6, in the 300s and 280s; 59, pl. 7, in the 220s; 61, 63, pl. 7–8, in the 190s; 64, pl. 8, in the 180s; for a Corinthian-helmeted Athenian on an atypical silver pentobol from the 260s, see 10–11, 22, pl. 3. 72 See Killen, Parasema 195–199. 73 G. Cohen, The Hellenistic Settlements in Europe, the Islands, and Asia Minor (Berkeley 1995) 84, for bibliography. 74 Though on occasion the reverse has the striding Athena wielding a thunderbolt: Mørkholm, Early Hellenistic Coinage 135–137, cf. 163–164. 234 PHILIP V AND LYSIMACHEIA: AN OATH IN GOLD ————— Greek, Roman, and Byzantine Studies 62 (2022) 203–238 first pi and the left hasta of the second are shortened so as not to touch the thunderbolt. In other words, the apparent intrusion of the thunderbolt into ΦΙΛΙΠΠΟΝ was either planned, or, I suspect, the thunderbolt was fashioned on the tablet prior to the in- scribing of the text, which would, then, be the case also with the repoussé helmeted Athena, perhaps marking the hand of two different artisans. Be that as it may, associating the thunderbolt with Philip is visually inescapable, which allows Athena to over- see the oath as the protector of Lysimacheia, for which Philip is obliged to wield his arms, the marble shield found at Lysima- cheia serving to signify that role as well. Conclusion The stone inscription found at Dion of the alliance between Philip and Lysimacheia, in spite of its fragmentary state, still speaks distinctly as a public monument of the promise to aid and be aided by an ally if attacked by a hostile force. The stone shield found at Lysimacheia, sporting Philip’s name and the image of Heracles’ club, documents distinctly this same moment, those very few years between 202/1 and winter 199/8, when the Lysimacheians saw in their city a marble monument of their alliance with Philip, presumably inscribed with the full text of the treaty, and backed up by the Macedonian troops then safeguarding their city. If that stone monument, of which the shield was some part, was erected by the Lysimacheians as a commemoration of this alliance with Philip, might they have made for Philip a token of it as well, in gold? At 2 to 3 mm. thick and weighing 40 gm, perhaps this was enough to impress a king.75 But I cannot cite an example of such tokens. Perhaps, though it lacks mounting holes, such as are found on every inscribed bronze plaque that still preserves at least one corner, its pedimental shape suggests that it was fixed in a decorative frame and put on display in the palace at Pella, part of a trophy 75 The Orphic lamellae found in burials are often called “leaves”; the Cretan lamellae studied in Y. Tzifopoulos, ‘Paradise’ Earned: The Bacchic-Orphic Gold Lamellae of Crete (Washington 2010), are less than 1/10 of a mm. BRAD L. COOK 235 ————— Greek, Roman, and Byzantine Studies 62 (2022) 203–238 wall of military and diplomatic victories. After all, Dion was some 40 km distant to the south, and his ancient enemies, the Aetolians, in 219 B.C., had dared to march on Dion and destroy everything at the sacred shrine “and even toppled the statues of the kings” (Polyb. 4.62.1–3); and to forge his alliance with the Lysimacheians Philip had just persuaded them to break their alliance with the Aetolians, expelling an Aetolian general from Lysimacheia in the process (Polyb. 18.3.11). But whether it was some sort of desktop token or mounted on a display wall in the palace, a reminder of the full alliance inscription at Dion, imagine the directions given to the artisan, whether Lysima- cheian or Macedonian: the king’s tablet needs to be of gold, royally weighty, but small enough that we can hand it to him, which means that the text needs to be legible but an epitome of what the alliance states. The result is that the start of the oath that Philip took, “I will be an ally of the Lysimacheians accord- ing to the agreed terms,” to which is added “and if anyone goes to war against Philip or against the Lysimacheians, the Ly- simacheians will aid Philip and Philip the Lysimacheians,” a compressed form of the standard, full phraseology but one that is precisely like what is found on the inscription of the Macedonian-Chalcidian treaty of ca. 393/2, at least on the copy at Olynthus. In the absence of ancient comparanda that might offer the slightest basis for either of these imaginary functions of the gold tablet, all the research on discrete features of the artifact may seem pointless. The fact, for instance, that the Amyntas- Chalcidian inscription shows that in 202 B.C. a goldsmith could most certainly have epitomized Philip’s oath that way, could just as easily support a different imagining. What if the recipient was, instead, a modern collector of such artifacts, one who owned already a palm-sized inscription on bronze and another inscription on a rolled sheet of silver, one who had crisscrossed Greece and Thrace almost as much as Philip V, who would, then, be surely delighted to turn his duo into a trio, crowned with 236 PHILIP V AND LYSIMACHEIA: AN OATH IN GOLD ————— Greek, Roman, and Byzantine Studies 62 (2022) 203–238 gold, viz., David M. Robinson. The features of the gold tablet must lead to something like one of these two scenarios, and, as with all such artifacts bereft of their archaeological context, if we knew where and in what place this artifact was found, we might be able to consider further this question. In the frustrating ab- sence of such information, my view is that the former of these two scenarios, that the gold tablet was made in antiquity to commemorate Philip V’s promise of aid and protection, is less unlikely, and that it, like the stone shield at Lysimacheia, broken from its original place on some larger monument, still stands to this day as a witness to the importance of alliances in antiquity.76 June, 2022 Department of Classics University of Mississippi blcook@olemiss.edu 76 Perhaps the next question someone else may wish to investigate is whether the gold fillet with an odd inscription in repoussé, in the portion of Robinson’s collection that went to Harvard, inv. 1960.663, can clarify anything about the gold artifact that stayed in Oxford; see https:// harvardartmuseums.org/collections/object/290173 (accessed 9 June 2022). It seems probable, at least, that it is like the gold tablet insofar as all informa- tion about its acquisition died with David Robinson, but it does differ from the gold tablet in that Robinson did not ask anyone to publish it. I owe many thanks to Kent Rigsby at the start and end of this project, to the University of Mississippi Museum and especially Melanie Antonelli, the collections manager, to my Department and the College of Liberal Arts for research support, and to the NEH and the ASCSA for a five-month research grant in 2018, where the superb library as well as input from a number of scholars in Athens influenced my research, though I alone am responsible for the conclusions here presented; in the writing of this article I have benefited throughout from the advice of Kerri J. Hame and, in the final stage, from the referees for the journal. BRAD L. COOK 237 ————— Greek, Roman, and Byzantine Studies 62 (2022) 203–238 Figure 1: Gold tablet, University of Mississippi Museum inv. 77.3.2273 238 PHILIP V AND LYSIMACHEIA: AN OATH IN GOLD ————— Greek, Roman, and Byzantine Studies 62 (2022) 203–238 Figure 2: X-ray of gold tablet, inv. 77.3.2273, courtesy of University Health Services Figure 3: start of line 4, gold tablet, inv. 77.3.2273