Dexipp ————— Greek, Roman, and Byzantine Studies 62 (2022) 437–493 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/ Facing the Plague and the Goths: A New Passage from the Scythica Vindobonensia (Codex Vindobonensis hist. gr. 73, fol. 192r, lines 13–30) Gunther Martin and Jana Grusková HIS ARTICLE presents for the first time 18 new manu- script lines (averaging 34 letters per line), 55% of one page, of the Scythica Vindobonensia, new historical frag- ments dealing with invasions of “Scythians” into the Roman Empire around the middle of the third century A.D. These fragments have been preserved in a palimpsest in the Austrian National Library in Vienna1 and are universally assigned to the Skythika of the contemporary historian Dexippus of Athens. The new text is part of a narrative of a Scythian invasion into Thrace, Macedonia, and Greece which has been dated to ca. 254 or the early 260s.2 It immediately precedes a section of two manuscript pages that has already been published (fols. 192v+ 193r).3 In addition to unknown information about the Gothic 1 For the discovery see J. Grusková, Untersuchungen zu den griechischen Palimpsesten der Österreichischen Nationalbibliothek (Vienna 2010) 50–53. See also Acknowledgments below. 2 On parallel sources, which seem to describe the same invasion, see below, especially the section on Date. 3 See G. Martin and J. Grusková, “‘Dexippus Vindobonensis (?)’. Ein neues Handschriftenfragment zum sog. Herulereinfall der Jahre 267/268,” WS 127 (2014) 101–120; the authors retracted this dating already in Grusková and Martin, Tyche 29 (2014) 38–39 (see n.6 below), and J. Grusková and G. Martin, “Rückkehr zu den Thermopylen: Die Fortsetzung einer Erfolgsgeschichte in den neuen Fragmenten Dexipps von Athen,” in T 438 FACING THE PLAGUE AND THE GOTHS ————— Greek, Roman, and Byzantine Studies 62 (2022) 437–493 campaign, the new lines offer a contemporary glimpse into catastrophic events in Rome at that time, in particular the so- called ‘Plague of Cyprian.ʼ4 Decipherment Thoroughly washed off the parchment and largely covered by later writing, the text of the Scythica Vindobonensia alias Dexippus Vindobonensis is very hard to retrieve. Its decipherment has been a labour-intensive and time-consuming task, which is ongoing. This process relies on a combination of in-depth palaeographical analysis and careful philological scrutiny. It has been essential to cooperate closely with specialists in the digital recovery (i.e. imaging and image processing) of written artefacts who by applying state-of-the-art methods have rendered the remnants of the erased writing visible. Since the discovery two research projects have focused on further deciphering, examining, and editing the Vienna frag- ments (see Acknowledgments below). The final goal has been their comprehensive critical edition (editio princeps) and analysis.5 The first stage of this work drew on special images of the eight pages of the palimpsest collected by methods of multispectral capture and advanced image processing. The results were preliminary transcriptions (accompanied by studies on the text) of six pages: fol. 195rv (= fr. Ia+Ib), fol. 194rv (= fr. IIa+IIb)―both belong- ing to an invasion of 250–2516―and the already mentioned ___ Ch. Schubert et al. (eds.), Das dritte Jahrhundert. Kontinuitäten, Brüche, Übergänge. Ergebnisse der Tagung der Mommsen-Gesellschaft am 21.-22.11.2014 an der Ber- gischen Universität Wuppertal (Stuttgart 2017) 267–281, at 269–270. 4 See the bibliography on the Scythica Vindobonensia in F. Mitthof, G. Martin, and J. Grusková (eds.), Empire in Crisis: Gothic Invasions and Roman Historiography (Vienna 2020 [Tyche Suppl. 12]) 565–570 (Anhang III). 5 The first volume of the edition, containing fol. 195rv (fr. I) and fol. 194rv (fr. II), will be published by the authors of this article in due course. 6 See G. Martin and J. Grusková, “‘Scythica Vindobonensia’ by Dexip- pus(?): New Fragments on Decius’ Gothic Wars,” GRBS 54 (2014) 728–754; J. Grusková and G. Martin, “Ein neues Textstück aus den ‘Scythica Vindo- GUNTHER MARTIN AND JANA GRUSKOVÁ 439 ————— Greek, Roman, and Byzantine Studies 62 (2022) 437–493 fols. 192v+193r (= fr. IIIb+IIIc).7 The text on fol. 192r (= fr. IIIa) and fol. 193v (= fr. IIId) was of such poor legibility that additional methods of recovery were required. This article now offers the first fruits of the decipherment and the philological and historical analysis of fol. 192r lines 13–30.8 More than ten years of work have been necessary to arrive at the current degree of legibility of these 18 new lines. A large number and variety of special images have been created. We have thoroughly examined each one in order to detect all sur- viving information on the underlying writing. One can rec- ognize about twenty characters of the undertext with the naked eye ( fig. 1). The multispectral imaging (MSI) and special image processing has rendered accessible, scattered over the page, about 100 identifiable characters and a number of faint strokes that probably belong to the undertext ( fig. 2). By arranging a high-tech experiment of fast-scanning XRF element mapping at a synchrotron we collected scans of individual elements, such as iron, calcium, etc. present in the Vienna folios. It turned out that the original ink has been almost completely removed and left no more than ‘footprints’ in the parchment. Years of processing and examining every single area followed ( fig. 3–4). Some general limitations of XRF scans affected the work. Re- solution is low in this kind of measuring experiment (this contrasts with the very high resolution of MSI images) and information on the text from the other side of the parchment ___ bonensia’ zu den Ereignissen nach der Eroberung von Philippopolis,” Tyche 29 (2014) 29–43; “Zum Angriff der Goten unter Kniva auf eine thrakische Stadt (Scythica Vindobonensia, f. 195v),” Tyche 30 (2015) 35–53; “Neugelesener Text im Wiener Dexipp-Palimpsest (Scythica Vindobonensia, f. 195v, Z. 6–10) mit Hilfe der Röntgenfluoreszenzanalyse,” ZPE 204 (2017) 40–46. 7 On the division of the text into three fragments see Martin and Grusková, in Empire in Crisis 543–548 (Anhang I), at 544. 8 The first 12 lines of the undertext on fol. 192r are still illegible (except for a few letters). On the digital recovery see below with n.144. 440 FACING THE PLAGUE AND THE GOTHS ————— Greek, Roman, and Byzantine Studies 62 (2022) 437–493 leaf often appears and contaminates the relevant data. For fur- ther details on the digital recovery see Acknowledgments below. Deciphering individual characters and words on the basis of the faint, fragmentary remnants of the original writing proved most demanding, sometimes impossible. The difficulties were aggravated by palaeographical peculiarities: On fols. 192rv+ 193rv, the eleventh-century scribe used many cursive elements and considerable variations in the form (minuscule/majuscule), the shape and the size of one and the same character (see fig. 3– 4). He moved his hand obviously faster than on fols. 195rv and 194rv.9 Countless sessions of repeated examination of the images have been required to detect traces of relevant infor- mation and make progress in deciphering the faint and partly concealed characters. Text The transcription and the edition given below represent the current state of the decipherment. Some letters are still in- visible, others are too faint to be identified with certainty. A dot under a letter indicates that the letter is still doubtful. Given the remaining uncertainties, the wording, and hence the meaning, may change in the future. It will require a great deal of ad- ditional effort to make further progress in recovering and de- ciphering the text. Considering the unique evidence contained in the new passage, we have decided to make our preliminary results available at this stage despite some degree of uncer- tainty. We thereby hope to initiate a discussion on the new pas- 9 On the manuscript see G. De Gregorio, E. Gamillscheg, J. Grusková, O. Kresten, G. Martin, B. Mondrain, and N. Wilson, “Palaeographical and Codicological Remarks on the Vienna Dexippus Palimpsest,” in Empire in Crisis 5–13; see also J. Grusková and G. De Gregorio, “Neue paläo- graphische Einblicke in einige palimpsestierte Handschriften aus den griechischen Beständen der Österreichischen Nationalbibliothek,” in C. Rapp et al. (eds.), New Light on Old Manuscripts: The Sinai Palimpsests and Other Advances in Palimpsest Studies (Vienna [in press]). GUNTHER MARTIN AND JANA GRUSKOVÁ 441 ————— Greek, Roman, and Byzantine Studies 62 (2022) 437–493 sage, the results of which will be included in the critical edition of the Scythica Vindobonensia. Transcription In the following work-in-progress transcription, the orthogra- phy and the punctuation10 of the manuscript are faithfully reproduced.11 As for accents and breathing marks, only those are reproduced which have been reliably identified. Abbrevia- tions (for -ος in lines 18 and 19, and probably -ης in 28) are resolved in round brackets. We have separated words through- out the text. The scribe apparently intended to do so but was not consistent. A hyphen has been set (by us) where a word is divided between two lines. Asterisks replace unreadable letters. Details are discussed in the commentary. Cod. Vind. hist. gr. 73, fol. 192r, lines 13–30 of the lower text: 13 τ̣ω̣ αληθεῖ εἰκ̣αζετο̣· Ῥω̣µ̣αῖοι δε ⁎⁎⁎⁎⁎⁎⁎⁎ 14 Κι̣ν̣ήσει τοῦ ἄστεος· π̣ρος τε ἀ̣λλοις θεων ιερο̣ις 15 ἱκέτευσαν· κατὰ τὸ τῆς σιβύ̣λλης µ̣α̣ν̣τε̣υ̣µ̣α̣ 16 καὶ ἔθυσαν διῒ σωτῆρι· κ̣αὶ ἥρα· και̣ ⁎⁎⁎⁎⁎ 17 ποσειδῶνι γης κινητορι· ἥ̣ τε νοσος πολλη̣ τε̣ 18 καὶ ἀδιήγητ(ος) ἐν τούτω ἠρε̣τ̣ο̣·̣ ἐς τε την ρω- 19 µην τα πλεῖστα τῆς ἑλλάδ(ος) ἐγκατασκήψασα·̣ 20 βραχύταται µὲν γὰρ ο̣υτως και̣ εν τῶ πρὸ το̣ῦ̣ 21 α̣ι̣ δι̣ακωχαί· τοτε δὲ ἐπὶ πολλῆ τῆ αἰσθήσει ἰ- 22 σ̣χυ̣σεν· ἐλέχθη γε τοι̣ ἐφ ἡ̣µέρα̣·̣ ἑκάστη̣ των̣ 23 θνησκ̣ο̣ντων· πεντακισχιλίους ἢ̣ κ̣αὶ̣ προσω̣ 24 τ̣ο̣υ̣ς̣ υ̣πογραφεας ἐν τ̣η̣ ῥώµη̣ ἀ̣(ex ὑ̣)πογ̣ρ̣α̣ψ̣α̣ι̣ τ̣ε̣ - 25 ⁎⁎⁎⁎⁎⁎⁎⁎⁎⁎⁎·̣ ⁎⁎δ⁎ δη̣ τρυγητοῦ ὥρα αρκτου- 10 The Byzantine system of punctuation is not identical with the modern one. 11 The legibility of the left margin is insufficient to verify if horizontal strokes with a dot above and below have been inserted here as on other pages of the manuscript. 442 FACING THE PLAGUE AND THE GOTHS ————— Greek, Roman, and Byzantine Studies 62 (2022) 437–493 26 ρ̣ο̣υ̣ σ̣υ̣ν̣θέο̣ντος·̣ σκυθῶν ο̣ι̣ δὴ γοῦθοι κ̣ε- 27 κλη̣µ̣ε̣ν̣ο̣ι̣ ἡγουµένου γουθουρικου σφῶν ὃς 28 ηρχε̣ τῆς πάσης στρατιᾶ̣ς̣. ἄρτι εκ τ̣ης φυγ(ῆ̣ς̣) 29 κ̣ατακληθεὶς ναυσὶ πορευθέντες ὑπὲρ̣ τοῦ 30 ἴστρου τὸν πόρον; εἰσέβαλλον εἴς τε θρά-|| Edition and translation In the following preliminary edition, accentuation, breathing marks, and punctuation have been normalized. Abbreviations have been resolved. The first letter of proper nouns has been capitalized. Since the Vienna manuscript is the codex unicus for this text, all the letters we have added (including the iota mutum, which the eleventh-century scribe never wrote) are in pointed brackets ⟨ ⟩. In lines 17, 21, and 25, we have emended assumed scribal errors, providing the spelling of the codex in round brackets (“… Cod.”), in smaller print. In lines 13, 16, and 25, asterisks replace unreadable letters. The lines of the manuscript page have been numbered for convenience of reference. Further details are described in the commentary. In the translation, renderings and additions based on un- certain readings have been printed in italics. For the invisible text in lines 13 and 25 we offer hypothetical supplements in round brackets and in italics; in line 16, we have preferred not to suggest any translation and used three dots instead. Scythica Vindobonensia, fr. IIIa (fol. 192r) 13–30:12 Ῥω̣µ̣αῖοι δὲ ⁎⁎⁎⁎⁎⁎⁎⁎ |14 κι̣ν̣ήσει τοῦ ἄστεος π̣ρός τε ἄ̣λλοις θεῶν ἱερο̣ῖς |15 ἱκέτευσαν κατὰ τὸ τῆς Σιβύ̣λλης µ̣ά̣ν̣τε̣υ̣µ̣α̣ |16 καὶ ἔθυσαν Διῒ σωτῆρι κ̣αὶ Ἥρα⟨ι⟩ καὶ̣ ⁎⁎⁎⁎⁎ |17 Ποσειδῶνι γῆς κινητῆρι (κινητορι Cod.). ἥ̣ τε νόσος πολλή̣ τε̣ |18 καὶ ἀδιήγητος ἐν τούτω⟨ι⟩ ἤ⟨ι⟩ρε̣τ̣ο̣, ἔς τε τὴν Ῥώ|19µην ⟨καὶ⟩ τὰ πλεῖστα τῆς Ἑλλάδος 12 The first words of line 13 on fol. 192r τ̣ω̣ αληθεῖ εἰκ̣αζετο̣ (see the transcription) obviously belong to the previous sentence. GUNTHER MARTIN AND JANA GRUSKOVÁ 443 ————— Greek, Roman, and Byzantine Studies 62 (2022) 437–493 ἐγκατασκήψασα. |20 βραχύταται µὲν γὰρ ο̣ὕτως καὶ̣ ἐν τῶ⟨ι⟩ πρὸ το̣ῦ̣ |21 α̣ἱ̣ δι̣ακωχαί, τότε δὲ ἐπὶ πολὺ (πολλῆ Cod.) τῆ⟨ι⟩ αἰσθήσει ἴ|22σ̣χυ̣σεν· ἐλέχθη γέ τοι̣ ἐφ’ ἡ̣µέρα̣⟨ι⟩ ἑκάστη̣⟨ι⟩ τῶν̣ |23 θνη⟨ι⟩σκ̣ό̣ντων πεντακισχιλίους ἢ̣ κ̣αὶ̣ πρόσω̣ |24 τ̣ο̣ὺ̣ς̣ ὑ̣πογραφέας ἐν τ̣ῆ̣⟨ι⟩ Ῥώµη̣⟨ι⟩ ἀ̣πογ̣ρ̣ά̣ψ̣α̣ι̣ τ̣ε̣ |25⁎⁎⁎⁎⁎⁎⁎⁎⁎⁎⁎. ⁎⁎δ⁎ δὴ̣ τρυγήτου (τρυγητοῦ Cod.) ὥρα⟨ι⟩ Ἀρκτού|26ρ̣ο̣υ̣ σ̣υ̣ν̣θέο̣ντος, Σκυθῶν ο̣ἱ̣ δὴ Γοῦθοι κ̣ε|27κλη̣µ̣έ̣ν̣ο̣ι̣, ἡγουµένου Γουθουρίκου σφῶν, ὃς |28 ἦρχε̣ τῆς πάσης στρατιᾶ̣ς̣ ἄρτι ἐκ τ̣ῆς φυγῆ̣ς̣ |29 κ̣ατακληθείς, ναυσὶ πορευθέντες ὑπὲρ̣ τοῦ |30 Ἴστρου τὸν πόρον εἰσέβαλλον εἴς τε Θρά⟨ι⟩||(fol. 192v)κην καὶ Μακεδονίαν καὶ ἐληίζοντο τὴν αὐτό|2θι̣ γ̣ῆν σύµπασαν.13 The Romans, (upon this) (com)motion of the city, asked for mercy at other temples of the gods, in accordance with the oracle of the Sibyl, and sacrificed to Zeus the Saviour, to Hera, … to Poseidon, Mover of the Earth. And at this time the disease grew severe and indescribable, having struck Rome ⟨and⟩ most of Greece. For, while the inter- missions had been very brief in this way also in the past, at that time it [scil. the disease] became greatly overbearing because of how it manifested itself. At least it was reported that the hypographeis in Rome registered 5000 or even more dead (who succumbed to this disease) every day. And, at the time of the vintage, with which the rising of Arcturus coincides, those of the Scythians who are called Goths ―led by Gouthourikos, who commanded the entire army, having just been recalled from his exile―took the way across the Istros in ships, invaded Thra||(fol. 192v)ce and Macedonia and ravaged all the land there. Commentary The text forms part of the beginning of a narrative unit that deals with a new invasion of “Scythians” into the territory of the Roman Empire. It can be divided into three parts; the last one continues in lines 1–12 of fol. 192v: 13 For the continuation see Appendix I below. 444 FACING THE PLAGUE AND THE GOTHS ————— Greek, Roman, and Byzantine Studies 62 (2022) 437–493 (A) fol. 192r (fr. IIIa) lines 13–17 (Ῥω̣µ̣αῖοι δὲ … κινητῆρι); (B) fol. 192r (fr. IIIa) lines 17–25 (ἥ̣ τε νόσος … ἀ̣πογ̣ρ̣ά̣ψ̣α̣ι̣ τ̣ε̣ |⁎⁎⁎⁎⁎⁎⁎⁎⁎⁎⁎); (C) fol. 192r (fr. IIIa) lines 25–30 + fol. 192v (fr. IIIb) lines 1– 2 (⁎⁎δ⁎ δὴ̣ τρυγήτου ὥρα⟨ι⟩ … γ̣ῆν σύµπασαν). In A and B the author briefly describes the situation in which Rome found herself when the new invasion started (192r 13– 25). Then, in C, he first clarifies the chronology (192r 25–26) and succinctly introduces the invaders and their commander (192r 26–29). What follows is a very concise report of the first actions of the Goths on Roman territory: They cross the Istros, i.e. the Danube (192r 29–30), and invade and plunder the en- tire (provinces of) Thrace and Macedonia (192v 1–2). In the subsequent text (fr. IIIb+IIIc), after fleetingly mention- ing an unsuccessful siege of Thessalonica (192v 2–7), the author announces the intention of the Goths to advance on Athens and Achaia (192v 7–12) because of the riches stored in sanctu- aries there. In line 13 of fol. 192v a new section begins, and the attention turns to the Greeks. After news of the Scythians’ advance has reached them, the Greeks take defensive measures against the invaders at Thermopylae, including the fortification of the pass. A Roman commander Marianus, an Athenian Phi- lostratus, and a Boeotarch Dexippus are named as the generals elected to supervise the forces. The beginning of an address by Marianus to the troops follows.14 14 See Martin and Grusková, WS 127 (2014) 101–120, and Grusková and Martin, in Das dritte Jahrhundert 267–281, at 268–270, including important corrections; see also, e.g., C. P. Jones “Further Fragments of Dexippus,” at https://www.academia.edu/11913736/Further_Dexippus_online_ (April 2015), updated (July 2016) by “Further Fragments of Dexippus (2),” https://www.academia.edu/26199041/Further_Dexippus_2_ (both last ac- cessed 25 Nov. 2022); Ch. Mallan and C. Davenport, “Dexippus and the Gothic Invasions: Interpreting the New Vienna Fragment,” JRS 105 (2015) 203–226; I. Piso, “Bemerkungen zu Dexippos Vindobonensis (I),” GFA 18 (2015) 199–215; J. McInerney, Dexippus (FGrHist 100 Revised), Brill’s New GUNTHER MARTIN AND JANA GRUSKOVÁ 445 ————— Greek, Roman, and Byzantine Studies 62 (2022) 437–493 The text deciphered so far breaks off with the end of fol. 193r (fr. IIIc) in the middle of Marianus’ speech. The hitherto re- covered parts of the following page, fol. 193v (fr. IIId), are still too fragmentary to permit any conclusions about its content, except that the speech continued. The attention that Dexippus pays to and the way he describes the defensive measures taken against the invaders at Thermopylae, including Marianus’ speech, suggest that a victory over Gouthourikos and his Gothic army followed.15 One parallel account, in the Historia Augusta (Gall. 5.2–6.1), reports that the Goths were defeated in Achaia and then withdrew (6.1 pugnatum est in Ach⟨a⟩ia Mariano duce contra eosdem Gothos, unde victi per Ach⟨a⟩eos recesserunt).16 Other sources―Zosimus (1.29.2–3), George Syncellus (466.1–7 Moss- hammer), and Zonaras (12.23, III 139.26–140.1 Dindorf)― which seem to describe events from the same invasion, further mention that the Athenians refortified their city and the Peloponnesians built a wall across the Isthmus. Syncellus also notes that the invaders left with much plunder (466.7 οἱ δὲ Σκύθαι µετὰ πολλῶν λαφύρων εἰς τὰ ἴδια ἦλθον). Three texts are of particular interest because of their obvious inter- and intratextual relationship to the new passage. These are discussed more broadly in the section on Inter- and intra- textual relationships, but details are cited throughout the com- mentary. To facilitate reference, the second text in question is quoted in full in Appendix II, the first and the third in that section. The three texts are: 1) Passages in Thucydides’ History in which he refers to or de- scribes the plague that struck Athens in 430–426 B.C.: 1.23.3 ___ Jacoby online―Second Edition (2019). 15 See Grusková and Martin, in Das dritte Jahrhundert 269–270; Piso, GFA 18 (2015) 215; for a more elaborate scenario see Mallan and Davenport, JRS 105 (2015) 221. 16 We follow the edition of E. Hohl, Scriptores Historiae Augustae II3 (Leipzig 1971); see Appendix II below. 446 FACING THE PLAGUE AND THE GOTHS ————— Greek, Roman, and Byzantine Studies 62 (2022) 437–493 (an announcement of the plague in the chapter on the reasons and the sufferings of the war), 2.47.3–4 (the beginning of a detailed account of the plague), and 3.87.1–4 (its second out- break). 2) The aforementioned section of the HA Life of Gallienus (5.2– 6.1). 3) A passage in the recently identified prooemium of Dexippus’ Skythika, i.e. the work to which the Scythica Vindobonensia (in all likelihood) belong. The prooemium has survived among the excerpts from Dexippus in De sententiis in the anthology com- missioned by Constantine VII (945–959) and transmitted in the palimpsest Vat. gr. 73.17 The passage in question covers fol. 54r (p. 107) lines 10–15. (A) Cod. Vind. hist. gr. 73, fol. 192r (fr. IIIa) lines 13–17 (the lemmata have been taken from the edition) 13–14 Ῥω̣µ̣αῖοι δὲ ⁎⁎⁎⁎⁎⁎⁎⁎ | κι̣ν̣ήσει τοῦ ἄστεος (“The Romans, (upon this) (com)motion of the city”): The scribe marks a new section: An initial letter kappa in κι̣ν̣ήσει is set at the beginning of line 14, the first complete line of the section (see fig. 2 and 4). This kappa is written in a majuscule form and is approximately three times larger than the kappas in the text. It extends beyond the ruled frame into the left margin. A different ink, apparently of a reddish hue, was used. Similar initials occur in the manuscript elsewhere, signalling a new section or paragraph, e.g. in fol. 192v (fr. IIIb) 13, where the narrative shifts from one group (the Scythians) to another (the Greeks) (see Appendix I ).18 17 The prooemium was deciphered and identified by András Németh of the Biblioteca Apostolica Vaticana thanks to new advances in the digital re- covery of the palimpsest: A. Németh, “Dexippus in the Excerpta Constan- tiniana Revisited: The Preface to Dexippus’ Scythica,” in Empire in Crisis 111– 134. 18 In line 13 there is a blank space of about two characters before Ῥω̣µ̣αῖοι, which may also (but need not) have served to signal the start of the GUNTHER MARTIN AND JANA GRUSKOVÁ 447 ————— Greek, Roman, and Byzantine Studies 62 (2022) 437–493 The visible letters δε after Ῥω̣µ̣αῖοι in line 13 could belong to the particle δέ (there is a faint stroke above the epsilon that may have been a gravis accent). Since the preceding lines (192r 1– 12) are still illegible, it remains unknown whether this δέ (if correctly identified) indicates a shift to a new topic or if the same topic is continued and δέ is connective. Ῥω̣µ̣αῖοι is the subject of the sentence in lines 13–17. This is the only use of the term “Romans” so far in the narrative in the Scythica Vindobonensia.19 The context, especially the mention of the ἄστυ, may suggest that these are the inhabitants of the city of Rome. The meaning could, however, also be wider (e.g. the population of the Roman Empire). The characters at the end of line 13 are illegible. As the lines on this page do not all have the same length (see fig. 3; the difference is equal to the space of 1–3 letters), one could think of various conjectures: e.g. ἐν/ἐπὶ ταύτη⟨ι⟩ τῆ⟨ι⟩ (“upon this”) or ἐν/ἐπὶ αὐτῆ⟨ι⟩ τῆ⟨ι⟩ (“upon just this”).20 τοῦ ἄστεος: The scribe uses the termination -ος for the gen- itive of τὸ ἄστυ, as in fol. 195r (fr. Ia) 17,21 a form common in imperial Greek prose.22 The mention of the Romans and the Sibylline Books indicates that ἄστυ is a shorthand for the city of ___ new section. 19 The only other occurrence is in a speech, Marianus’ address to the “Greek” army in fol. 193r (fr. IIIc) 17–22, and it refers to events in 191 B.C.: οἱ πρόγονοι ὑµῶν … Ῥωµαίοις̣ ἄρχο̣υ̣σιν̣ ἤδη συνιστάµενο̣ι̣. 20 The visible remnants do not favour ἐν τῆ⟨ι⟩δε τῆ⟨ι⟩. 21 Τhe genitive sg. of τὸ ἄστυ occurs also in fol. 195v (fr. Ib) 4, but the penultimate character has not yet been rendered visible to the degree that one can decide, which of the two forms was written there. 22 It is not used in the ‘old’ fragments of Dexippus; the spelling ἄστεως occurs twice in fr.22 Martin (Skythika) (= fr.28 Mecella, FGrHist 100 F 25) and once in fr.27 Martin (Skythika) (= fr.33 Mecella, FGrHist F 29). Fragment numbers are given according to F. Jacoby, FGrHist; G. Martin, Dexipp von Athen (Tübingen 2006); L. Mecella, Dexippo di Atene (Tivoli 2013). Cf. also McInerney in Brill’s New Jacoby online. 448 FACING THE PLAGUE AND THE GOTHS ————— Greek, Roman, and Byzantine Studies 62 (2022) 437–493 Rome (like urbs).23 The word (ἡ) κίνησις can mean “motion/movement” in a political sense, but can also refer to a “motion/shaking” of the ground, i.e. an earthquake.24 In this instance, the latter mean- ing is supported by the information that the Romans sacrificed to “Poseidon, Mover of the Earth” (see the commentary on line 17). The corresponding report in HA Gall. 5–6 mentions both political upheaval and a severe earthquake (5.2 inter tot bellicas clades etiam terrae motus gravissimus fuit; 5.4 mota est et Roma, mota Libya; 5.6 terrae motus). We prefer to withhold the final decision on the meaning of this word and give a neutral translation until the preceding text (192r 1–12) has been recovered. In any case, it is obvious that what we read in the Scythica Vindobonensia has a wording very close to HA Gall. 5. Irrespective of the meaning of κίνησις, the sacrifices to “Poseidon, Mover of the Earth” strongly suggest that the Scythica Vindobonensia reported one or more earthquakes that struck the Romans, in line with the HA. Nicholas Ambraseys emphasises that it is not inconceivable that the HA synthesised several separate earthquakes, which took place within a few months or even years of each other.25 For Asia Minor, a catastrophic earthquake has been attested archaeologically in 23 This meaning of ἄστυ is common in imperial literature, cf. e.g. Cass. Dio 40.48.1 τοιαύτης οὖν τότε τῆς ἐν τῷ ἄστει καταστάσεως οὔσης… 24 Both meanings of the noun or the verb κινέω are attested in Thu- cydides: war is a human motion (1.1.2 κίνησις … αὕτη µεγίστη … ἐγένετο of the Peloponnesian war; 3.82.1 πᾶν … τὸ Ἑλληνικὸν ἐκινήθη; cf. 3.75.3), earthquakes a non-human motion (2.8.3 Δῆλος ἐκινήθη …, πρότερον οὔπω σεισθεῖσα). See S. N. Jaffe, Thucydides on the Outbreak of War: Character and Contest (Oxford 2017) 68 with n.29. 25 N. Ambraseys, Earthquakes in the Mediterranean and Middle East: A Multi- disciplinary Study of Seismicity up to 1900 (Cambridge 2009) 137–139, esp. 138. See also E. Guidoboni, A. Comastri, and G. Traina, Catalogue of Ancient Earthquakes in the Mediterranean Area up to the 10th Century (Rome 1994) 241– 245. GUNTHER MARTIN AND JANA GRUSKOVÁ 449 ————— Greek, Roman, and Byzantine Studies 62 (2022) 437–493 Ephesus.26 Cyrene in Libya seems to have been destroyed by an earthquake about this time, for large parts of the city were rebuilt during the reign of Claudius II (268–270).27 In the ‘old’ fragments of Dexippus, i.e. those known before the discovery of the Scythica Vindobonensia (FGrHist 100/Martin 2006/Mecella 2013), no mention of an earthquake is made. However, in the recently deciphered prooemium to the Skythika, Dexippus men- tions earthquakes (Vat. gr. 73, fol. 54r [p. 107] 13 σεισ̣µοῖς; quoted below in context) among those sufferings which oc- curred during the war and which he will describe along with the military events. 14–17 π̣ρός τε ἄ̣λλοις θεῶν ἱερο̣ῖς | ἱκέτευσαν κατὰ τὸ τῆς Σιβύ̣λλης µ̣ά̣ν̣τε̣υ̣µ̣α̣ | καὶ ἔθυσαν Διῒ σωτῆρι κ̣αὶ Ἥρα⟨ι⟩ καὶ̣ ⁎⁎⁎⁎⁎ | Ποσειδῶνι γῆς κινητῆρι (κινητορι Cod.) (“asked for mercy at other temples of the gods, in accordance with the oracle of the Sibyl, and sacrificed to Zeus the Saviour, to Hera, … to Poseidon, Mover of the Earth”): In κατὰ τὸ τῆς Σιβύ̣λλης µ̣ά̣ν̣τε̣υ̣µ̣α̣, the reference is obviously to the Libri Sibyllini. Romans of the Republican era turned to them traditionally after disquieting omens, in wars or in times of extraordinary natural calamities such as earthquakes, floods, plague, and the like.28 Such events were understood as prodigia, i.e. as an expression of divine anger, which made atonement necessary. In accordance with a Senate resolution, the secret books, written in Greek, were consulted by the XVviri sacris faciundis to restore the pax deorum (cf. Varro Rust. 1.1.3; Liv. 10.47.6–7). After the original Sibylline Books were destroyed in a fire in 83 B.C. (Dion. Hal. Ant.Rom. 4.62), a reconstructed ver- 26 S. Ladstätter and A. Pülz, “Ephesus in the Late Roman and Early Byzantine Period,” in A. Poulter (ed.), The Transition to Late Antiquity on the Danube and Beyond (Oxford 2007) 391–434, at 394–397. 27 See Ambraseys, Earthquakes 138 (with further bibliography). 28 In connection with earthquakes the Books were consulted e.g. in 174 B.C. (Liv. 41.28.2), with plagues in 348 B.C. (7.27.1) and 293 B.C. (10.47.6). 450 FACING THE PLAGUE AND THE GOTHS ————— Greek, Roman, and Byzantine Studies 62 (2022) 437–493 sion was reportedly consulted down to the fourth century A.D., though information is scarce:29 we hear of four consultations in the third century A.D. A dubious one is reported in Aur. Vict. Caes. 34.3 and Epit. de Caes. 34.3 (under Claudius II [268–270] in a context different from ours). For the other three the only source is the HA. According to the HA, a consultation took place during the reign of Gordian III (238–244) after a violent earthquake.30 The Books were allegedly consulted again, as HA Aurel. 18.4–5 reports, during an invasion by the Marcomanni under Aurelian (270–275). The third attestation in the HA is at Gall. 5.5, pax igitur deum quaesita inspectis Sibyll⟨a⟩e libris factu⟨m⟩que Iovi Salutari, ut praeceptum fuerat, sacrificium (“the favour of the gods was sought by consulting the Sibylline Books and sacrifices to Iuppiter Salu- taris were performed in accordance with their order”)―part of the passage that correlates closely throughout with fr. III of the Vienna palimpsest. The events are dated by the HA to “the consulship of Gallienus and Faustianus,”31 i.e. 262, and the context is a mixture of disasters, comprising defeats in war and natural catastrophes.32 It seems obvious that HA Gall. 5.5 and the Scythica Vindo- bonensia are speaking of the same consultation. In addition, it is 29 Cf. H. W. Parke, Sibyls and Sibylline Prophecy in Classical Antiquity (London 1988) 206–212, 215; D. S. Potter, Prophecy and History in the Crisis of the Roman Empire: A Historical Commentary on the Thirteenth Sibylline Oracle (Oxford 1990) 109–114. The text of the Books was not itself publicly available, but the Senate proclaimed (in Latin) the instructions for the specific situation after their consultation, cf. Lactant. Div.Inst. 1.6.13. 30 HA Gord. 26.1–2: Fuit terrae motus eo usque gravis imperante Gordiano, ut civitates etiam terrae hiatu cum populis deperirent. ob quae sacrificia per totam urbem totumque orbem terrarum ingentia celebrata sunt. et Cordus quidem dicit inspectis libris Sibyllinis celebratisque omnibus quae illic iussa videbantur mundanum malum esse sedatum. 31 On the form “Faustianus” see n.132 below. 32 On the uncertain date of these events see on Date below. GUNTHER MARTIN AND JANA GRUSKOVÁ 451 ————— Greek, Roman, and Byzantine Studies 62 (2022) 437–493 commonly assumed that Dexippus (presumably the Chronika) was the source from which the HA drew (for more see on Inter- and intratextual relationships below, under “b”).33 In the two instances in which the consultation was prompted by natural disasters―HA Gord. 26.1–2 and Gall. 5.5―the Libri ordered expiatory rites. Again, these seem to be the only at- testations of such rites34 after the reign of Nero (Tac. Ann. 15.23.2).35 The description that the Scythica Vindobonensia give matches the accounts from republican times, e.g. Liv. 27.11.6 (ea prodigia hostiis maioribus procurata, et supplicatio circa omnia pul- vinaria); cf. 36.21.9 (supplicatio in triduum decreta est et ut quadraginta hostiis maioribus praetor, quibus dis ei videretur, sacrificaret). According to these examples, it was common practice to have, on the one hand, supplicationes at all temples in Rome36 (cf. ἱκέτευσαν in line 15) and, on the other hand, sacrifices of animals to specific gods (cf. ἔθυσαν in line 16). The fact that both consultations of the Libri and expiatory rites in Rome had been attested only in the HA―often dispar- aged as a historical source―has led to their historicity being questioned.37 Now the Scythica Vindobonensia back up the ac- 33 It is likely that the consultation in HA Gord. 26.1–2 was also taken from Dexippus. The consultation under Aurelian, however, cannot have been reported in the Chronika, which ended with the reign of Claudius in 269. It may have been mentioned in the Skythika, but the HA probably did not use this work. Cf. the commentary by F. Paschoud, Histoire Auguste V.1 (Paris 1996) 116–118, who assumes that this invasion by the Marcomani is the same as that by the Iuthungi in Dexippus fr.28 Martin (= fr.34 Mecella, FGrHist F 6). 34 Supplications in general are attested during this period, but these are thanksgiving rites, cf. e.g. R. Selinger, Die Religionspolitik des Kaisers Decius (Frankfurt am Main 1994) 52–76. 35 The evidence has last been collected by L. Halkin, La supplication d’action de grâces chez les Romains (Paris 1953) 10. 36 Cf. K. Latte, Römische Religionsgeschichte (Munich 1960) 245–246. 37 Cf. e.g. Parke, in Sibyls 211; F. Kolb, Untersuchungen zur Historia Augusta 452 FACING THE PLAGUE AND THE GOTHS ————— Greek, Roman, and Byzantine Studies 62 (2022) 437–493 count of HA Gall.38 While one may be inclined to give Dexip- pus more credit than the HA, we still cannot rule out that Dexippus might be imitating descriptions of similar scenarios which he found in his predecessors’ works, such as Livy and other annalists.39 If, on the other hand, the account in the Scythica Vindobonensia is historically accurate, the lack of at- testations for the Libri and expiatory supplications may be a consequence of the general scarcity of historiographical sources after Tacitus. 14–16 π̣ρός τε ἄ̣λλοις θεῶν ἱερο̣ῖς | ἱκέτευσαν … | καὶ ἔθυσαν: The formulation seems to imitate Thuc. 2.47.4 πρὸς ἱεροῖς ἱκέτευσαν ἢ µαντείοις καὶ τοῖς τοιούτοις ἐχρήσαντο. The expression π̣ρός τε ἄ̣λλοις … καὶ seems to follow the pat- tern of ἄλλος τε … καὶ. Since it links the two main clauses, τε is moved forward to the second place of the sentence. The an- tithesis is twofold: it juxtaposes supplication and blood sacrifice, and it emphasises the three deities mentioned by name over all the others. Extant parallels (such as si deo si deae and similar ex- pressions, cf. e.g. Gell. 2.28.3) may suggest that ἄ̣λλοις θεῶν ἱερο̣ῖς comprises all gods with temples in Rome, including Zeus, Hera, and Poseidon; cf. e.g. Liv. 27.11.6 ea prodigia hostiis mai- oribus procurata, et supplicatio circa omnia pulvinaria. 15 κατὰ τὸ τῆς Σιβύ̣λλης µ̣ά̣ν̣τε̣υ̣µ̣α̣: The letters at the end of line 15, after σιβύ̣λλης, are, apart from τ, insufficiently visible for a reliable identification. The word µάντευµα, the oracle or prophecy of the Sibyl, would ___ (Bonn 1987) 138–140. The only other mention is highly dubious, as stated above. 38 Further support for continued consultation of the Books may be lent by the attestation of the XVviri sacris faciundis down to the fourth century: cf. N. Hächler, Kontinuität und Wandel des Senatorenstandes im Zeitalter der Soldatenkaiser (Leiden 2019) 720. 39 It has been established that Dexippus archaizes, e.g. in his siege de- scriptions: Martin, Dexipp von Athen 226–229. GUNTHER MARTIN AND JANA GRUSKOVÁ 453 ————— Greek, Roman, and Byzantine Studies 62 (2022) 437–493 match the faint traces of the characters and fit in with the article τὸ and the genitive τῆς Σιβύ̣λλης. No close Greek parallel has emerged for this wording, but cf. e.g. Plin. HN 18.286 ex oraculis Sibyllae and Amm. Marc. 30.4.11 Sibyllae oraculorum interpretes. The supplication and the sacrifices are commonly or- dered together in Latin sources; hence, we take the expression κατὰ τὸ τῆς Σιβύ̣λλης µ̣ά̣ν̣τε̣υ̣µ̣α̣ in substance to refer to both ἱκέτευσαν and ἔθυσαν in lines 14–16; cf. HA Gall. 5.5 ut prae- ceptum fuerat (scil. a Sibylla). 16–17 καὶ ἔθυσαν Διῒ σωτῆρι κ̣αὶ Ἥρα⟨ι⟩ καὶ̣ ⁎⁎⁎⁎⁎ | Ποσειδῶνι γῆς κινητῆρι (κινητορι Cod.): Part of the text corresponds closely to HA Gall. 5.5 factu⟨m⟩que Iovi Salutari, ut praeceptum fuerat, sacrificium. The names of Zeus, Hera, and Poseidon have been deciphered with certainty. At the end of line 16 of the manuscript, after ἥρα· και̣, faint traces of about five characters follow. Those characters have not been rendered sufficiently visible for reliable decipherment. The first two could be a (minuscule) gamma followed by a (majuscule) eta.40 As for the following text, based on some visible strokes one could think of ·̣ κ̣α̣ι̣, but the area is too fuzzy to verify it. If γ̣η̣·̣ κ̣α̣ι̣ were to be correct, the (edited) reading would be κ̣αὶ Ἥρα⟨ι⟩ καὶ̣ Γ̣ῆ̣⟨ι⟩ κ̣α̣ὶ̣ | Ποσειδῶνι.41 However, given the poor legibility the end of line 16 must remain open for the moment. Ζεὺς Σωτήρ or his Latin equivalent Iuppiter Salutaris, i.e. Jupiter the Saviour or Healer, is an appropriate recipient of sacrifices after serious natural catastrophes. In Greece, worship 40 One might think of the epithet of Poseidon γηοῦχος, an exceedingly rare variant of γαιήοχος “earth-moving / earth-carrying”; cf. e.g. Hsch. γ 512 (ed. Latte/Cunningham) γηοῦχος· ὁ τὴν γῆν συνέχων (an interpolation from Lex.Rhet.); Hesychius also provides (γ 50) γαιήοχος· ὁ τὴν γῆν συνέχων. 41 Carlo M. Lucarini (per litteras) suggests αὐτῶ(ι). This attractive conjecture would suit the context―as it points to Poseidon’s being the most pertinent deity to pray to after an earthquake―but it does not correspond to the visible strokes. 454 FACING THE PLAGUE AND THE GOTHS ————— Greek, Roman, and Byzantine Studies 62 (2022) 437–493 of Ζεὺς Σωτήρ was common.42 Iuppiter Salutaris is occasionally referred to in Latin literary texts,43 and attested in some Latin inscriptions.44 Hera (line 16) follows Zeus, as she often does in Roman sacrifices.45 If Γ̣ῆ̣⟨ι⟩ (line 16) were to be correct, the Latin equivalent would be Tellus, the Roman earth-goddess. Poseidon is called upon as god of earthquakes, κινήτωρ, “Mover of the Earth” (line 17).46 The word κινήτωρ seems not to occur elsewhere. It could be explained as a variation on κινητήρ, analogous to γεννήτωρ, -ορος / γεννητήρ, -ῆρος. Given the lack of attestations of κινήτωρ so far, we have assumed a scribal error and adopted κινητῆρι in the preliminary edition. The latter word is attested in the TLG twice, both times in poetry and with γῆς, as epiclesis of Poseidon, god of earth- quakes (Pind. Isthm. 3/4.37 ὁ κινητὴρ δὲ γᾶς; Hymn.Hom. 22.2 γαίης κινητῆρα), and synonymous with ἐννοσίγαιος, ἐνοσίχθων, or similar. It has been pointed out that the Latin god Neptune, a water deity responsible for rivers, fresh water sources, and all 42 On his cult in classical Athens see R. Parker, Athenian Religion (Oxford 1996) 238–241; he was worshipped throughout Greece, cf. M. P. Nilsson, Geschichte der griechischen Religion I3 (Munich 1967) 414–416. In connection with an earthquake, Ζεὺς Σωτήρ seems to appear in a private inscription by an inhabitant of Byblos, thanking for his survival: L. Robert, “Documents d’Asie Mineure,” BCH 102 (1978) 395–543, at 399: [Διὶ Σω]τῆρι Ἀπολλό- δωρος Νίκωνο(ς) ἀπὸ σεισµοῦ διασωθεὶς ἀνέθηκεν. 43 E.g. Cic. Fin. 3.66; Plin. HN 2.34. 44 E.g. AE 1980, 793.1; CIL III 6456 [ILS 3025]. 45 Cf. e.g. U. Ehmig, “Der ‘Erdbebengott Neptun’ und die ‘un- bestimmten Erdbebengötter’ in lateinischen Inschriften,” in J. Borsch et al. (eds.), Erdbeben in der Antike (Tübingen 2016) 37–59, at 45. Along with Jupiter, Juno is one of few gods to whom sacrifices were made during the Ludi Saeculares. The last of these Secular Games were celebrated in 248, under Philip the Arab (244–249), on the occasion of Rome’s first mil- lennium. 46 See Ehmig, in Erdbeben 37–59, and I. Mylonopoulos, “Poseidon und Neptun. Zwei Götter – Zwei polytheistische Systeme,” Polifemo 5 (2005) 240–254. GUNTHER MARTIN AND JANA GRUSKOVÁ 455 ————— Greek, Roman, and Byzantine Studies 62 (2022) 437–493 kinds of water, does not appear as linked to earthquakes and seaquakes.47 His importance in Roman cult does not generally equal that of Poseidon in Greece.48 Therefore, Ποσειδῶν γῆς κινητήρ in line 17 most probably shows Greek influence. To sum up, the combination of the three gods suits the occasion. It reflects the expiatory nature of the rite to be performed in order to bring back good fortune after the current catastrophes. The passage in lines 14–17 reveals substantial Greek influ- ence. The choice of gods appears to have reflected the Greek religious background of the Sibyl. If Dexippus’ narrative is historically accurate, the three names were translated from the Greek of the Sibylline Books (which were written in Greek hexameters) into Latin and from there retranslated into Greek by Dexippus or his source(s). If, however, the information is fictitious, Dexippus or his source(s) have chosen the gods that seemed most appropriate to them, without much concern for the Roman pantheon. (B) Cod. Vind. hist. gr. 73, fol. 192r (fr. IIIa) lines 17–25 17–19 ἥ̣ τε νόσος πολλή̣ τε̣ | καὶ ἀδιήγητος ἐν τούτω⟨ι⟩ ἤ⟨ι⟩ρε̣τ̣ο̣, ἔς τε τὴν Ῥώ|µην ⟨καὶ⟩ τὰ πλεῖστα τῆς Ἑλλάδος ἐγκατασκήψασα (“And at this time the disease grew severe and indescribable, having struck Rome ⟨and⟩ most of Greece”): Dexippus announces in the prooemium to the Skythika that he will deal with sufferings caused by a “disease,” which he characterises as “pestilential” (Vat. gr. 73, fol. 54r [p. 107] 12– 47 Inscriptions do not link Neptune with sacrifices after seismic events; cf. Ehmig, in Erdbeben 43. However, Neptune was included in the first lectisternium in 399 B.C., also ordered by the Sibyl: Dion. Hal. Ant.Rom. 12.9, Liv. 5.13.6. See G. Wissowa, Religion und Kultus der Römer2 (Munich 1912) 225–227; Latte, Religionsgeschichte 131 with n.4. According to Dionysius of Halicarnassus (2.31.2), some people identified the Latin water god Consus (Κῶνσος) with Poseidon Seisichthon, “Earthshaker.” 48 Cf. e.g. Mylonopoulos, Polifemo 5 (2005) 240–254, esp. 240–241. 456 FACING THE PLAGUE AND THE GOTHS ————— Greek, Roman, and Byzantine Studies 62 (2022) 437–493 13 ἐν νόσωι τ̣ῆι λοι|µώδει).49 The consequences of the νόσος mentioned on fol. 192v 17–25 of the palimpsest leave no doubt that the word is used in the same specific sense (cf. LSJ s.v. A). Moreover, the parallel account in the HA has the unequivocal term pestilentia: Gall. 5.5 nam et pestilentia tanta extiterat vel Romae vel in Achaicis urbibus, and 5.6 ex diversis partibus pestilentia orbem Romanum vastaret. The passage on the plague (lines 17–25) concludes the de- scription of the situation in the Empire at the time of the new Scythian campaign. It is connected with the passage on the κίνησις (ending in line 17) by the particle τε. While it is unclear exactly how the wider section is structured as a whole (since its first part is missing), the plague as an extraordinary event with an enormous death toll would be a fitting climax. The brevity of the passage and the perfunctory introduction of the plague (ἥ̣ τε νόσος) as well as the formulation ἐν τῶ⟨ι⟩ πρὸ το̣ῦ̣ (line 20) indicate that the plague is familiar to the readers and this is not its first appearance in the narrative (i.e. outside of the pro- oemium).50 Fol. 192r 17–25 now seem to be depicting the peak of this disease.51 The Scythica Vindobonensia are obviously referring to the so- called Plague of Cyprian, which ravaged the Roman Empire in the third quarter of the third century. The information the author gives and its contextualization agree with our previous knowledge about this epidemic.52 The most detailed evidence is 49 See on Inter- and intratextual relationships below, under “c”; also Gunther Martin (in preparation). Cf. Thuc. 1.23.3 καὶ ἡ οὐχ ἥκιστα βλάψασα καὶ µέρος τι φθείρασα ἡ λοιµώδης νόσος. 50 One might hope to find confirmation of an earlier report (or reports) in the Skythika from mentions of the plague in the HA prior to Gall. 5.5. However, such is precluded by the ‘lacuna’ from ca. A.D. 244 to 253. Cf. O. Desbordes and St. Ratti, Histoire Auguste IV.2 (Paris 2000) XIX–XXVIII. 51 See on Date below. 52 The ‘Plague of Cyprianʼ has received new attention in recent years, see K. Harper, “Pandemics and Passages to Late Antiquity: Rethinking the GUNTHER MARTIN AND JANA GRUSKOVÁ 457 ————— Greek, Roman, and Byzantine Studies 62 (2022) 437–493 provided by Zosimus. He first mentions the plague in the reign of Gallus (and Volusianus) (251–253), interestingly in connec- tion with a Scythian invasion (Zos. 1.26.1–2), then in con- nection with Valerian’s confrontation with Shapur (1.36.1) and again soon after the capture of Valerian in 260 (1.37.3).53 Ac- cording to him, a plague “fell upon the cities, such as had never occurred before in all of history” (1.37.3 λοιµὸς ἐπιβρίσας ταῖς πόλεσιν, οἷος οὔπω πρότερον ἐν παντὶ τῷ χρόνῳ συνέβη).54 It is debated whether and to what extent this account is based on Dexippus.55 Christopher Jones and Sabine Huebner in their recent contributions agree that the ‘Cyprianic plagueʼ arrived in the Roman Empire in the early 250’s―it may have broken out soon after Decius’ death (251)―and petered out in the late 260’s (see on Date below).56 The adjective πολλή̣ in connection with a plague could be interpreted either as “severe” or as “widespread.” Both find a counterpart in HA Gall. (5.5 tanta, 5.6 ex diversis partibus). The ___ Plague of c. 249–70 Described by Cyprian,” JRA 28 (2015) 223–260; K. Harper, The Fate of Rome: Climate, Disease, and the End of an Empire (Princeton 2017) 136–144; C. P. Jones, “Dexippus and the Third-Century Plague,” in Empire in Crisis 159–164; S. R. Huebner, “The ‘Plague of Cyprian’: A Revised View of the Origin and Spread of a 3rd-c. CE Pan- demic,” JRA 34 (2021) 151–174. 53 See Jones, in Empire in Crisis 161–162. Cf. the commentary in F. Paschoud, Zosime: Histoire Nouvelle I2 (Paris 2000) 156–158. 54 Transl. Jones, in Empire in Crisis 162. 55 See e.g. F. Paschoud, “L’Histoire Auguste et Dexippe,” in G. Bona- mente et al. (eds.), Historiae Augustae Colloquium Parisinum (Macerata 1991) 217–269; Jones, in Empire in Crisis 163. 56 See Jones, in Empire in Crisis 159–164, and Huebner, JRA 34 (2021) 156–163, against Harper, JRA 28 (2015) 227, who dates the earliest attesta- tion in Egypt to 249. According to a report by the sixth-century Evagrius, the third-century historian Philostratus stated that it lasted 15 years (HE 4.29, p.179.9–12 Bidez/Parmentier = Philostratus KFHist A3 fr.2), which is probably an approximation, see Jones, in Empire in Crisis 160. For further discussion see on Date below. 458 FACING THE PLAGUE AND THE GOTHS ————— Greek, Roman, and Byzantine Studies 62 (2022) 437–493 second adjective, ἀδιήγητος, could in this context be para- phrased as “which cannot be described because of the extreme degree of suffering.”57 For a combination of the two adjectives, cf. Eunap. FHG fr.4 (fr.5.1 Blockley, ap. Suda κ 391, s.v. Κα- ρῖνος, πολλῆς οὔσης καὶ ἀδιηγήτου τῆς φθορᾶς) and Eunap. VS 7.3.4 (πολλῶν καὶ ἀδιηγήτων ἐπικλυσθέντων κακῶν). See also on lines 18–19 below. For an independent contemporary view of the gravity of this plague, cf. e.g. Eus. HE 7.22.6. ἤ⟨ι⟩ρε̣τ̣ο̣: The first letter seems to be a small majuscule eta and there is a small stroke that could be a spiritus lenis (the accent is not visible); the second letter seems to be rho (rather than nu, the minuscule form of which is very similar to rho in this particular Greek minuscule script; see fig. 3–4). The next, faint traces could belong to τ̣ο̣ or ε̣τ̣ο̣ with epsilon and tau in ligature (as in line 15 in ἱκέτευσαν), but these readings are by no means certain. If τ̣ο̣ is correct, one could read ἦρτ̣ο̣, i.e. pluperfect med.(pass.) of αἴρω; if ε̣τ̣ο̣ is correct, one could read ἤ⟨ι⟩ρε̣τ̣ο̣, i.e. impf. med.(pass.) of αἴρω in the meaning of “to be raised/ increase” (cf. ἡ δύναµις ᾔρετο Thuc. 1.118.2) with a predicative use of the adjectives πολλή̣ and ἀδιήγητος (cf. Dion. Hal. Ant. Rom. 7.65.1 ὁ δῆµος ἤρθη µέγας, Plut. Cat.Mai. 23.3 ἡ πόλις ἤρθη µεγίστη).58 18–19 ἔς τε τὴν Ῥώ|µην ⟨καὶ⟩ τὰ πλεῖστα τῆς Ἑλλάδος ἐγκατα- σκήψασα: A free space of 1–2 letters precedes the first epsilon. At the beginning of line 19, the mu extends beyond the ruled frame. 57 Suda α 482 (ed. Adler) ἀδιήγητον· τὸ µὴ δυνάµενον δι᾿ ὑπερβολὴν κακοῦ διηγηθῆναι. Cf. Xen. Cyr. 8.7.22; Plut. Fab.Max. 11.6.3; Philostorgius KFHist E7 fr.7.1 ἄρρητά τε καὶ ἀδιήγητα πάθη. 58 From the fact that Thucydides uses ἤρξατο for the plague (2.47.3, 2.48.1, 2.54.5), it might be tempting to look for possible readings (or pos- sible restorations) among the forms of ἄρχοµαι. The most suited would be ἤρχετο, i.e. ἤρ⟨χ⟩ε̣τ̣ο̣. However, neither the durative or iterative aspect of the imperfect with ἄρχοµαι nor the semantics seems to fit. We are grateful to Nigel Wilson for sharing his considerations on this matter. GUNTHER MARTIN AND JANA GRUSKOVÁ 459 ————— Greek, Roman, and Byzantine Studies 62 (2022) 437–493 Hence, the scribe may have intended to indicate a new para- graph, on the plague, by using a letter of the usual size and shape as an initial, similar to fol. 195v (fr. Ib) 3. If so, one would, however, have to assume that the scribe forgot to set the initial in line 18 (the first complete line of the paragraph), realized his mistake, and corrected it a line later. Otherwise, this detail may be one of the meaningless irregular palaeo- graphical features that often occur in this manuscript. The characters are legible without ambiguity. Τhe first four letters can be read as either ἔς τε or ἔστε. If ἔς τε is to be read, the preposition ἐς belongs to ἐγκατασκήπτω (see below) and the participle refers to both Rome and Greece. In this case, a con- nector needs to be added after Ῥώµην to correspond with τε, i.e. either ἔς τε τὴν Ῥώ|µην ⟨καὶ⟩ τὰ πλεῖστα τῆς Ἑλλάδος ἐγκατα- σκήψασα or ἔς τε τὴν Ῥώ|µην τά ⟨τε⟩ πλεῖστα etc., “having struck both Rome and Greece” (the latter, a suggestion by Carlo M. Lucarini, is palaeographically easier, but no parallel for the use of τε … τε to connect words has been found in Dexippus’ frag- ments yet). This is apparently what the HA read and it puts the clear focus on the adjectives πολλή̣ and ἀδιήγητος. If, by contrast, Dexippus meant ἔστε, a preposition of space, “up to,” then ἔστε τὴν Ῥώ|µην belongs to the previous text as an adverbial phrase meaning “as far as Rome.” Another prepo- sition ⟨ἐς⟩ then needs to be added to ἐγκατασκήψασα (see below), as pointed out to us by Nigel Wilson: the omission could easily have been caused by homoeoarchon, i.e. ἥ̣ τε νόσος πολλή̣ τε̣ | καὶ ἀδιήγητος ἐν τούτω⟨ι⟩ ἤ⟨ι⟩ρε̣τ̣ο̣ ἔστε τὴν Ῥώ|µην, ⟨ἐς⟩ τὰ πλεῖστα τῆς Ἑλλάδος ἐγκατασκήψασα, “And at this time the disease grew severe and indescribable as far as Rome, having struck most of Greece.” This wording would suit the tenor of the entire pas- sage, lines 13–25, in that it is focused on the situation in Rome. The HA (Gall. 5.5. nam et pestilentia tanta extiterat vel Romae vel in Achaicis urbibus), however, does not seem to have understood the passage in this way. Moreover, the main clause would convey two competing new pieces of information―the severity of the illness and its geographical spread (“up to Rome”)―without a 460 FACING THE PLAGUE AND THE GOTHS ————— Greek, Roman, and Byzantine Studies 62 (2022) 437–493 clear focus. Above all, ἔστε without following ἐπί is not securely attested in literary texts. Arr. Ind. 2.2 and 41.8 are the only examples (against more than 80 instances with ἐπί, all in Arrian) and have hence been doubted.59 ἐγκατασκήπτω, “fall upon” of epidemics (cf. LSJ s.v. A intr.), is used by Thuc. 2.47.3 with πολλαχόσε (ἡ νόσος … καὶ πρότερον πολλαχόσε ἐγκατασκῆψαι, “that the plague had already struck widely elsewhere”), a passage which Dexippus seems to imitate here (see on Inter- and intratextual relationships below, under “a”). This verb is constructed with dative (never with accusative) or preposition, usually εἰς (+acc.); cf. Cass. Dio 49.15.5 ἐπειδὴ κε- ραυνὸς ἐς αὐτὸν ἐγκατέσκηψε and Ael. Arist. 19.258 (Keil/Behr) νέφος δὲ … ἐγκατέσκηψεν εἰς τὰς τῶν βαρβάρων ναῦς. 20–22: an explanation is provided for why the plague is being considered as πολλή and ἀδιήγητος. The particle µέν in line 20 is correlated with δέ in line 21. 20–21 βραχύταται µὲν γὰρ ο̣ὕτως καὶ̣ ἐν τῶ⟨ι⟩ πρὸ το̣ῦ̣ | α̣ἱ̣ δι̣ακωχαί (“For, while the intermissions had been very brief in this way also in the past”): In the word δι̣ακωχαί the scribe wrote alpha above iota (which is now concealed by the upper text), connecting the last stroke of this alpha with the first vertical stroke of the following kappa, which is therefore higher than usual (see fig. 3–4). The ακ in the word Μακεδονίαν ten lines later (fol. 192v 1) is written in a sim- ilar way.60 The reading διακωχή is a variant of διοκωχή (“inter- mission/cessation”) attested in Thucydides’ description of the Athenian plague (3.87.1 τοῦ δ’ ἐπιγιγνοµένου χειµῶνος ἡ νόσος τὸ 59 Arr. Ind. 2.2 παρατείνει ἔστε τὴν πρὸς ἕω θάλασσαν and 41.8 ἔστε Βαβυλῶνα, both in an Ionizing text; in his edition, Hercher has proposed ἔστε ⟨ἐπί⟩ in both instances. Other attestations of ἔστε alone are in dialect inscriptions, e.g. IK Priene 132.113, 166, 169; IG VII 3170.13. 60 See fig. 5 in J. Grusková, G. Martin, O. Kresten, and F. Mitthof, “Images of the Vienna Dexippus Palimpsest,” in Empire in Crisis 549–564 (Anhang II), at 555: line 1. GUNTHER MARTIN AND JANA GRUSKOVÁ 461 ————— Greek, Roman, and Byzantine Studies 62 (2022) 437–493 δεύτερον ἐπέπεσε τοῖς Ἀθηναίοις, ἐκλιποῦσα µὲν οὐδένα χρόνον τὸ παντάπασιν, ἐγένετο δέ τις ὅµως διοκωχή). It is mentioned e.g. in Hsch. δ 1075 διακωχή· ὁ µεταξὺ χρόνος, καὶ διάστηµα χρόνου and in Suda δ 601 διακωχή: διάλειψις. Cf. schol. Thuc. 3.87.1c (p.653.62 Kleinlogel/Alpers) (διακωχή:) διάλειψις, ἀναβολή (with apparatus). According to the edition of Giovanni Battista Alberti, the manuscripts of Thucydides uniformly transmit the variant διοκωχή.61 An emulation of Thucydides 3.87.1 is evi- dent, as διοκωχή (or διακωχή) is not found elsewhere before Cassius Dio62 (see on Inter- and intratextual relationships below, under “a”). ο̣ὕτως:63 The faint strokes at the beginning could correspond to omicron (although alpha cannot be completely excluded); the area is not sufficiently visible to decide. The sentence produces an antithesis between the earlier waves and the current one: all the waves have in common that the periods of reprieve (διακωχαί) have been very short, i.e. ο̣ὕτως καὶ̣ in the sense of “in this way also,” but the current wave is distinguished by its vehemence (τότε δὲ …). For ἐν τῷ πρὸ τοῦ scil. χρόνῳ cf. Thuc. 1.32.4 (where only a part of the manuscripts have χρόνῳ).64 21–22 τότε δὲ ἐπὶ πολὺ (πολλῆ Cod.) τῆ⟨ι⟩ αἰσθήσει ἴ|σ̣χυ̣σεν (“at that time it became greatly overbearing because of how it mani- 61 For διοκωχή cf. Ael. Dion. π 22 (ed. Erbse): παροκωχή· παροχὴ παρὰ Θουκυδίδῃ, ὡς ἡ ἀνοκωχή καὶ διοκωχή; the Lexicon of Photius π 432 (ed. Theodoridis): παροκωχή· παροχὴ παρὰ Θουκυδίδῃ, ὡς ἡ ἀνοκωχή καὶ διοκωχή, and Suda δ 1157 διοκωχή: ἀνακωχὴ χρόνου. ἐν µάχαις τισὶν αὐτοὺς νικήσας καὶ διοκωχὴν αἰτήσασιν ἔδωκεν. 62 Cass. Dio 39.47.2, 41.25.1, 47.27.2, 75.9.6, who uses the word in a different context and meaning (“armistice”). 63 In Greek prose οὕτως καί is well attested, whereas αὔτως/αὕτως καί without preceding ὡς δ’ is not. 64 Without χρόνῳ: Alberti 1972, Luschnat 1954; with χρόνῳ: Stuart Jones/ Powell 1900. The latter is more usual, e.g. Thuc. 1.103.2, 2.73.3; Lys. 12.2; Pl. Symp. 172D. 462 FACING THE PLAGUE AND THE GOTHS ————— Greek, Roman, and Byzantine Studies 62 (2022) 437–493 fested itself”): The subject changes: from the plural διακωχαί to a singular, apparently ἡ νόσος from line 17. The connection ἐπὶ πολλῆ⟨ι⟩ τῆ⟨ι⟩ αἰσθήσει seems obscure in connection with the verb ἰσχύω, and no suitable parallel is forthcoming.65 The solution may lie in emending the spelling of the manuscript: ἐπὶ πολλῆ⟨ι⟩ could have resulted from a simple hearing error from ἐπὶ πολύ, which is often used by Thucydides and well attested in other authors (cf. LSJ s.v. πολύς IV.4). The dative τῆ⟨ι⟩ αἰσθήσει can then be explained as one of cause. For αἴσθησις as “perception given” (LSJ I fin.) cf. e.g. Thuc. 2.50.2, 2.61.2. 22–25 ἐλέχθη γέ τοι̣ ἐφ’ ἡ̣µέρα̣⟨ι⟩ ἑκάστη̣⟨ι⟩ τῶν̣ | θνη⟨ι⟩σκ̣ό̣ντων πεντακισχιλίους ἢ̣ κ̣αὶ̣ πρόσω̣ | τ̣ο̣ὺ̣ς̣ ὑ̣πογραφέας ἐν τ̣ῆ̣⟨ι⟩ Ῥώµη̣⟨ι⟩ ἀ̣πογ̣ρ̣ά̣ψ̣α̣ι̣ τ̣ε̣|⁎⁎⁎⁎⁎⁎⁎⁎⁎⁎⁎ (“At least it was reported that the hypographeis in Rome registered 5000 or even more dead (who succumbed to this disease) every day”): 22 ἐλέχθη γέ τοι̣: This is one of the rare occasions when Dexippus, at least implicitly, talks about his sources of infor- mation (cf. Scyth. Vind. fol. 195r [fr. Ia] 17 ὡς ἐλέχθη). The source itself, however, remains unidentified. The strokes after ἐλέχθη and before ἐφ’ have been deciphered as γε τοι̣. The area above is not sufficiently visible to see if an accent was written there or not. The faint traces visible after το could belong to iota: τοι̣.66 If so, it is the first occurrence of this particle in the Scythica Vindobonensia (i.e. in the hitherto de- ciphered text). In the ‘old’ fragments of Dexippus, the only instance of τοι is S1b Martin = fr.3 Mecella = FGrHist F 34, a passage transmitted in the Suda without clear attribution.67 γέ 65 For ἐπί + αἰσθήσει cf. ἐπὶ στενοχωρίας αἰσθήσει Gal. De loc. aff. (VIII 284.6 Kühn). 66 Considering the traces and the space before ἐφ’, the reading τότ᾿ can be excluded. 67 Suda ε 2455 s.v. ἐπιµᾶλλον: […] καὶ αὖθις· ταῦτά τοι ἐπιµᾶλλον ἐξῆψεν αὐτῷ τὸ µῖσος. GUNTHER MARTIN AND JANA GRUSKOVÁ 463 ————— Greek, Roman, and Byzantine Studies 62 (2022) 437–493 τοι̣ would introduce an affirmation of the preceding assertion (cf. LSJ s.v. γε I.5).68 The following acc.-and-inf. construction τ̣ο̣ὺ̣ς̣ ὑ̣πογραφέας … ἀ̣πογ̣ρ̣ά̣ψ̣α̣ι̣ would constitute the subject of ἐλέχθη.69 22 ἐφ’ ἡ̣µέρα̣⟨ι⟩ ἑκάστη̣⟨ι⟩ (“every day”): This expression depicts the situation as even more disastrous than does HA Gall. 5.5 uno die, as it indicates that the high rate of fatalities occurred not only on one single day, but every day. 22–23 τῶν̣ | θνη⟨ι⟩σκ̣ό̣ντων πεντακισχιλίους ἢ̣ κ̣αὶ̣ πρόσω̣ (“5000 or even more dead”): θνησ and ντων have been deciphered with relative certainty, and the faint strokes after the sigma could belong to a kappa and an omicron. The genitive θνη⟨ι⟩σκ̣ό̣ντων could be explained as belonging to the following cardinal number, i.e. “5000 dead.”70 After πεντακισχιλίους, the letters α and προσ are legible, the letters ἢ̣ κ̣ and ὶ̣ are likely but uncertain. In the majority of images, the area after προσ is concealed to a great extent by the upper text or data from other layers. However, one of the processing methods rendered partly visible faint strokes that could belong to the undertext and be deciphered as an omega, with nothing more to follow in the line. Since at the beginning of line 24 the illegible space of 3–4 letters could correspond with τ̣ο̣ὺ̣ς̣, which one would expect before ὑ̣πογραφέας, this pos- sible (though uncertain) omega at the end of line 23 most likely goes with προσ to form πρόσω̣. The illegible area above the 68 For ἐλέχθη γέ τοι cf. e.g. Pl. Hp.mi. 367B7. 69 Cf. e.g. Arist. Eth.Nic. 1139a or Plut. Marc. 4.2.4. In the rather unlikely case that the stroke after το does not belong to a character of the undertext, το has to be interpreted as the article. For a substantivised acc.-and-inf. construction after ἐλέχθη cf. e.g. Ath.Pol. 21.2 ἐλέχθη τὸ µὴ φυλοκρινεῖν πρὸς τοὺς ἐξετάζειν τὰ γένη βουλοµένους. 70 For the use of the genitive with numerals cf. e.g. Hdt. 9.28.3 µετὰ δὲ τούτους ἵσταντο Κορινθίων πεντακισχίλιοι (“Next to these in the line were five thousand Corinthians”, transl. A. D. Godley). 464 FACING THE PLAGUE AND THE GOTHS ————— Greek, Roman, and Byzantine Studies 62 (2022) 437–493 letters does not allow any statement on accentuation. To sum up: the text after πεντακισχιλίους could be (preliminarily) edited as ἢ̣ κ̣αὶ̣ πρόσω̣.71 For πρόσω after a numeral―here 5000― meaning “more” cf. Clem. Alex. Strom. 1.21.102 γενεαὶ µὲν … διαριθµοῦνται, ἔτη δέ, ὡς ἔπος εἰπεῖν, τετρακόσια καὶ πρόσω; Epiph. Adv.Haeres. 78.8.2 (III 458.22–23 Holl) κατάγων ἡλικίαν περί που ὀγδοήκοντα ἐτῶν καὶ πρόσω ὁ ἀνήρ. Elsewhere in the Scythica Vindobonensia, που µάλιστα (“about, approximately”) is used to indicate the approximative character of large figures: cf. on the size of armies in fols. 195v (fr. Ib) 18–19, 194r (fr. IIa) 27–28, 194v (fr. IIb) 2–3. Here, by contrast, the author em- phasises that the figure he gives is the minimum, thereby intensifying the impression of its enormousness and making the disease appear particularly perilous. The text corresponds to HA Gall. 5.5 ut … quinque milia hominum pari morbo perirent. 24–25 τ̣ο̣ὺ̣ς̣ ὑ̣πογραφέας ἐν τ̣ῆ̣⟨ι⟩ Ῥώµη̣⟨ι⟩ ἀ̣πογ̣ρ̣ά̣ψ̣α̣ι̣ τ̣ε̣|⁎⁎⁎⁎⁎⁎⁎⁎⁎⁎⁎ (“that the hypographeis in Rome registered …”): On τ̣ο̣ὺ̣ς̣ see above. The first character of ὑ̣πογραφέας is very likely a broad upsilon which often appears before pi, e.g. in ὑπὲρ̣ in line 29. ἐν τ̣ῆ̣⟨ι⟩ Ῥώµη̣⟨ι⟩ (“in Rome”): the new text refers to Rome as the place where the deceased were registered by the ὑ̣πογραφεῖς. It does not tell us whether Rome was the place where these 5000 daily deaths occurred. The report of HA Gall. 5.5 nam et pestilentia tanta extiterat vel Romae vel in Achaicis urbibus, ut uno die quinque milia hominum pari morbo perirent seems to imply that the figure gives the total for the entire Empire or at least for Rome 71 The fragments of Dexippus otherwise only have the form πόρρω: cf. fr.22.4 Martin (= fr.28.4 Mecella, FGrHist F 25.4), fr.23.10 Martin (= fr.29.10 Mecella, FGrHist F 26.10), fr.30.2 Martin (= fr.36.2 Mecella, FGrHist F 7.2), also in the Scythica Vindobonensia in fol. 195r (fr. Ia) 5 and fol. 195v (fr. Ib) 5. The concurrence of both forms, i.e. πρόσω and πόρρω, in one work might seem peculiar, but it is not impossible (cf. e.g. Gal. Anat.Adm. II 714.18 and 715.4 Kühn; Arr. Anab. 3.28.8 and 3.28.9). GUNTHER MARTIN AND JANA GRUSKOVÁ 465 ————— Greek, Roman, and Byzantine Studies 62 (2022) 437–493 and Achaia combined.72 ἀ̣πογ̣ρ̣ά̣ψ̣α̣ι̣: the letter after Ῥώµη̣⟨ι⟩ and before the legible πο cannot be read with certainty. Judging from the visible traces, it nevertheless seems very likely that the scribe first wrote a wide upsilon with a spiritus asper (ὑπο-; this spiritus is visible), and then corrected it73 into an alpha with a spiritus lenis (ἀπο-; this spiritus is likewise visible) (see fig. 3–4). He may have been influenced by the previous ὑ̣πογραφέας (if correct). The faint traces of several characters after πο could belong to γ̣ρ̣ά̣ψ̣α̣ι̣, giving the inf. aor. act. ἀ̣πογ̣ρ̣ά̣ψ̣α̣ι̣. The meaning “register” (LSJ s.v. II.1) would suit the context well, with τῶν θνη⟨ι⟩σκόντων πεντακισχιλίους as a direct object. Interpreting ὑ̣πογραφέας is difficult firstly because the first character is not completely beyond doubt, secondly because it is uncertain which office and institution the text is referring to. In literary sources (e.g. Ar. Eq. 1256) and in documentary papyri (e.g. P.Oxy. VI 911 of A.D. 233 or 265) ὑπογραφεύς de- notes a person who signs for another. As an official title, the word appears in some inscriptions in Sicily (I.Akrai 6–8). John Lydus (6th cent.) in De magistratibus glosses it as scriba.74 As to the institution, the existence of a central register of deaths in Rome is a debated question. The closest we know of is the official register of burials in the temple of Libitina (Dion. Hal. Ant.Rom. 4.15.5, Suet. Nero 39.1) of which it is not clear whether it amounts to a complete systematic register of deaths.75 72 On the severe decline of the Empire’s population during the ‘Cyprianic plagueʼ see Harper, Fate 140–141. Similar counts of the deceased are re- ported in connection with other plagues: 30,000 in the autumn of 65 (Suet. Nero 39.1), 10,000 daily in 77 (George Sync. p.417.12; Eus. = Hieron. Chron. ab Abr. 2099 [188h Helm]). 73 For other corrections by the scribe himself see e.g. fol. 194r (fr. IIa) 7, 9. 74 Lydus Mag. 2.30 (128.24 Bandy): καὶ σκρίβαν µὲν ἐκείνῳ, ἀντὶ τοῦ ὑπογραφέα. 75 Hieron. Chron. ab Abr. 2099 (188h Helm) in efemeridem probably refers to the same document. Cf. C. Virlouvet, “Existait-il des registres de décès à 466 FACING THE PLAGUE AND THE GOTHS ————— Greek, Roman, and Byzantine Studies 62 (2022) 437–493 τ̣ε̣|⁎⁎⁎⁎⁎⁎⁎⁎⁎⁎⁎: The last letters in line 24 could be a tau and an epsilon. No other letters seem to have followed. The characters at the beginning of line 25 are invisible at present. From the context and the parallel evidence of HA Gall. 5.5 pari morbo perirent one might expect a word or words with the meaning “died” or “succumbed to this illness.” The traces are too meagre and uncertain to allow more than conjectures. Considering the syntax and the evidence of the HA, one would expect an infinitive or a participle belonging to the accusative πεντακισχιλίους: One may think of e.g. τελευτῆσαι, τεθνάναι, or τεθνεῶτας αὐτῆι (scil. τῆι νόσωι) or δι᾿ αὐτὴν (scil. τὴν νόσον). Round brackets in the translation indicate our hypothetical supplement. Alternatively, Markus Stein suggests that τ̣ε̣ after ἀ̣πογ̣ρ̣ά̣ψ̣α̣ι̣ could be a particle and indicate that καί with a similar expression of about nine characters followed (e.g. ἀριθµῆσαι). (C) Cod. Vind. hist. gr. 73, fol. 192r (fr. IIIa) lines 25–30 + fol. 192v (fr. IIIb) lines 1–2 25–26 ⁎⁎δ⁎ δὴ̣ τρυγήτου (τρυγητοῦ Cod.) ὥρα⟨ι⟩ Ἀρκτού|ρ̣ο̣υ̣ σ̣υ̣ν̣θέ- ο̣ντος (“And, at the time of the vintage, with which the rising of Arcturus coincides”): The ancient grammatical tradition distinguishes between the proparoxytonon ὁ τρύγητος (τρυγήτου) meaning (LSJ s.v. I.A.1) “gathering of fruits, vintage, harvest” or (I A 2) “time thereof, harvest or vintage,” and the oxytonon ὁ τρυγητός (τρυγητοῦ) ___ Rome au Ier siècle ap. J.-C.?” in C. Virlouvet, La Rome impériale. Démographie et logistique (Rome 1997) 77–88; D. G. Kyle, Spectacles of Death in Ancient Rome (London 1998) 180 n.77; S. Schrumpf, Bestattung und Bestattungswesen im Römi- schen Reich: Ablauf, soziale Dimension und ökonomische Bedeutung der Totenfürsorge im lateinischen Westen (Göttingen 2006) 235. Caution against interpreting these as references to systematic registration of the dead in the Roman Empire is advised by T. G. Parkin, Demography and Roman Society (Baltimore 1992) 37– 38. We owe important remarks and references on this topic to Fritz Mitthof. GUNTHER MARTIN AND JANA GRUSKOVÁ 467 ————— Greek, Roman, and Byzantine Studies 62 (2022) 437–493 meaning τὸ τρυγώµενον, “fruit gathered” (LSJ II).76 In the Vienna manuscript the eleventh-century scribe accented the genitive with a circumflex on the last syllable: τρυγητοῦ. The connection of the genitive with ἡ ὥρα, however, implies the first meaning. Since the transmitted accentuation in ancient works is generally considered to be a later addition, we have corrected the accent and written τρυγήτου in the edition, following the authority of Aelius Herodianus. ὥρα⟨ι⟩: considering the syntax of the sentence, it seems ob- vious that the transmitted form is to be understood as dative. At the beginning of the sentence one can recognise only two deltas with certainty. The traces after the second delta could correspond to a majuscule eta (the area above it is illegible), but the poor visibility makes verification impossible. The function of δὴ would be connective, indicating a transition to a new segment. τῆι δὲ δὴ or ἐν δὲ δὴ77 might seem plausible recon- structions, but owing to the very poor visibility, they cannot be verified. 25–26 Ἀρκτού|ρ̣ο̣υ̣ σ̣υ̣ν̣θέο̣ντος: the letters αρκτου at the end of line 25 have been deciphered with some degree of certainty. The beginning of line 26 is illegible: the faint traces could correspond to ρ̣ο̣υ̣, although ρ̣ω̣, i.e. dat. ρ̣ω̣⟨ι⟩, cannot be completely excluded. Considering the following σ̣υ̣ν̣θέο̣ντος, the genitive Ἀρκτούρ̣ο̣υ̣ seems more likely: in Greek latitudes, the heliacal (or morning) rising of Arcturus is in mid-September78 76 See Ps.-Arcadius p.216.9–10 Roussou = p.93.14–15 Schmidt), which is an Epitome of the Καθολικὴ προσῳδία of the grammarian Aelius Herodianus: τρύγητος (ὁ καιρὸς µονογενῶς, τρυγητός δὲ τὸ τρυγώµενον). 77 Cf. Paus. 8.28.2, Longus 3.1.2, and Sozom. HE 6.6.1 ἐν ὥρᾳ χειµῶνος. 78 In Greece Arcturus would begin to be visible as a morning star about the middle of September. For Thrace in Dexippus’ time it could have been about 21/22 September. See F. K. Ginzel, Handbuch der mathematischen und technischen Chronologie: das Zeitrechnungswesen der Völker II (Leipzig 1911) 521. 468 FACING THE PLAGUE AND THE GOTHS ————— Greek, Roman, and Byzantine Studies 62 (2022) 437–493 approximately and coincides with the vintage.79 σ̣υ̣ν̣θέο̣ντος: the last four letters as well as the theta and the epsilon have been deciphered with some degree of certainty. The faint remnants of the other characters seem to correspond to συν and ο. The use of συνθέω is similar to Pl. Leg. 844E1–2 πρὶν ἐλθεῖν τὴν ὥραν τὴν τοῦ τρυγᾶν Ἀρκτούρῳ σύνδροµον. The precise indication of time matches Dexippus’ obsession with exact chronology in the Chronika, as criticised by Eunapius: at FHG fr.1 (fr.1.6 Blockley) he gives the (possibly fictitious) example τὴν ἐν Σαλαµῖνι ναυµαχίαν ἐνίκων οἱ Ἕλληνες κυνὸς ἐπιτέλλοντος. This is the first time that we see a formulation of this type in Dexippus’ own text. However, such an indication of time is not unique in a historical text: cf. Thuc. 2.78.2 περὶ ἀρκτούρου ἐπιτολάς. Considering the redundance in the double indication of time (Arcturus and vintage), the parallel may, as Carlo M. Lucarini suggests, also be emulation of Thucydides in typically Dexippean style. 26–27 Σκυθῶν ο̣ἱ̣ δὴ Γοῦθοι κ̣ε|κλη̣µ̣έ̣ν̣ο̣ι̣ (“those of the Scythians who are called Goths”): The ethnics Σκυθῶν and Γοῦθοι are legible without ambiguity (see fig. 3). The spelling of the Greek name of the Goths with -οῦ- is not without parallel, and so is the single -θ-80 (cf. also lat. Gothi), although the latter is less frequent than -τθ- or -θθ-.81 The wording in fol. 192r (fr. IIIa) 26–27 clearly shows that the Scythica Vindobonensia consider the Γοῦθοι (Goths) as a sub- 79 Cf. O. Wenskus, Astronomische Zeitangaben von Homer bis Theophrast (Stutt- gart 1990) 176; e.g. Hes. Op. 609–611, Theophr. Hist.Pl. 5.1.2.4. 80 See M. Schönfeld, Wörterbuch der altgermanischen Personen- und Völkernamen nach der Überlieferung des klassischen Altertums (Heidelberg 1911) 120–123; M. H. Jellinek, Geschichte der gotischen Sprache (Berlin 1926) 3; W. Braune and F. Heidermanns, Gotische Grammatik20 (Tübingen 2004) 3. Cf. e.g. Strab. 7.1.3 Γούτωνας; Chron.Pasch. 46.12 Γόθους, 472.12 Γόθων, 508.6 Γόθοι (ed. Din- dorf, sic Vat. gr. 1941); Hsch. σ 737 (ed. Latte/Hansen) Γουθικόν. 81 E.g. I.SyriaAAES IIIA 223.1 (Inat, A.D. 208) µνηµεῖον Γούθθα, and SEG XX 324.7 (Persepolis, ante 272) Γούθθων τε καὶ Γερµανῶν ἐθνῶν. GUNTHER MARTIN AND JANA GRUSKOVÁ 469 ————— Greek, Roman, and Byzantine Studies 62 (2022) 437–493 group of the Scythians.82 At the start of the narrative of the new invasion Dexippus specifies the identity of the invaders. Only half a page later, he calls them Σκύθαι: fol. 192v (fr. IIIb) 13–14 ἐπεὶ δὲ̣ ἐ̣ς̣ τοὺς Ἕλληνας ἐξηγγέλθη ἡ ἔφοδος τ̣ῶν | Σκυθῶ̣ν̣. The same treatment of ethnics as in fr. III of the Scythica Vindobonensia can be found at the start of Syncellus’ report on the invasion of 250/1 (459.5ff. = Dexippus fr.17 Martin = fr.23 Mecella = FGrHist 100 F 22): Σκύθαι περαιωθέντες οἱ λεγόµενοι Γότθοι τὸν ῎Ιστρον ποταµὸν ἐπὶ Δεκίου πλεῖστοι τὴν ῾Ρωµαίων ἐπι- κράτειαν κατενέµοντο.83 Later in the same campaign Dexippus refers to the invaders only as “Scythians”: fol. 195r (fr. Ia) 20 ἀνέπεισε τοὺς Σκύθας, fol. 194r (fr. IIa) 17 ὁ τῶν Σκυθῶν ἄρχων, 19–20 οἱ Σκ̣ύθαι Κνίβαν µὲν ἐν λόγω⟨ι⟩ τῶ⟨ι⟩ ἀρίστω⟨ι⟩ | ἐπ̣ο̣ι̣οῦ̣ν̣το, and 25–26 ἐ̣ς̣ τ̣ὸ̣ | κ̣οινὸν Σκυθῶν84 (the source of Syncellus’ report, potentially down to the phrasing, is probably the Chronika).85 Hence, a pattern seems to emerge: at the start of a 82 Cf. e.g. Physiologus 7.2–4 (ed. Sbordone) … ἐν τοῖς Σκυθοῖς µέρεσιν, ἤτοι κατὰ βορρᾶν, ἔνθα εἰσὶν οἱ Γότθοι καὶ Δάνεις, (⟨δι⟩ὸ καὶ Σκύθας καλοῦσιν οἱ παλαιοὶ πάντας τοὺς κατοικοῦντας τὸ κλῖµα τὸ βόρειον). The same inter- pretation may be possible also for HA Gall. 6.2 (see below) or Philostorgius KFHist E7 fr.11.8.4 Ἀλλ’ ἐπὶ τούτοις πᾶσι καὶ Τριγίβιλδος, ἀνὴρ Σκύθης µὲν γένος τῶν νῦν ἐπικαλουµένων Γότθων (πλεῖστα γὰρ καὶ διάφορα τούτων ἐστὶν τῶν Σκυθῶν γένη) (ed. Bleckmann/Stein). Cf. Etym.Magn. 238 s.v. Γοῦτθος: Ὁ ἄρχων Σκυθῶν τῶν δὴ καλουµένων Γούτθων. Ἔοικε γὰρ ἀπὸ τοῦ ἡγεµόνος αὐ- τῶν κληθῆναι. We express our gratitude to the late Klaus Alpers for the following expertise: “Der Artikel Goutthos stammt aus Oros, und zwar aus Peri Ethnikon.” 83 Cf. Martin, Dexipp von Athen 105: “die Goten genannten Skythen”; Mecella, Dexippo di Atene 287: “alcuni tra gli Sciti, chiamati Goti.” 84 Cf. also recently S. Ghosh, Writing the Barbarian Past: Studies in Early Medieval Historical Narrative (Leiden 2016) 40: “The existing fragments of Dexippus speak almost always of Scythians (“Σκύθαι”); the one exception is a reference to Scythians who are spoken of as Goths attacking Histria in the reign of Decius (reigned 249–251) (Jacoby, ed. 1926: frag. 22 = Martin, ed. and trans. 2006: frag. 17: Σκύθαι […] oἱ λεγόµενοι Γότθοι). This might indicate that Dexippus distinguishes Goths as a sub-group of Scythians.” 85 There may even be a third example: HA Gall. 6.2 Scythae autem, hoc est 470 FACING THE PLAGUE AND THE GOTHS ————— Greek, Roman, and Byzantine Studies 62 (2022) 437–493 campaign Dexippus specifies the narrow ethnic affiliation of the invaders, while also classing them as belonging to the wider group he calls Scythians. Later in the narrative he simply calls them Scythians. As Dexippus introduces the same sub-group twice according to this pattern, he may have spoken about another sub-group of the Scythians in between. The new evidence of the Scythica Vindobonensia also raises a wider point about the conception of Dexippus’ Skythika. We now see that he has knowledge of and uses the name Goths. In his account the ethnic name Σκύθαι serves him as an umbrella term for a larger group of tribes.86 It follows that Dexippus is consciously describing the activities of heterogeneous groups of peoples, not a war between the Roman Empire and one ethnic group. It also proves his knowledge about the composition of the population north of the Roman frontier. 27–29 ἡγουµένου Γουθουρίκου σφῶν, ὃς | ἦρχε̣ τῆς πάσης στρατιᾶ̣ς̣ ἄρτι ἐκ τ̣ῆς φυγῆ̣ς̣ | κ̣ατακληθείς (“led by Gouthourikos, who commanded the entire army, having just been recalled from his exile”): Each character of the proper name Γουθουρίκου can be identified without doubt (see fig. 3); the gamma is written in a majuscule form (as often on this page, see e.g. Γοῦθοι in line 26). The ending -ου of this non-Greek name implies that the ___ pars Gothorum, Asiam vastabant, “i.e. that part which consists of Goths” or “namely their Gothic part.” It could, however, also mean “i.e. a part of the Goths,” in which case Bruno Bleckmann suggests to us that this may be a deliberate inversion on the part of the HA. Again, we are at the start of a campaign and the author of the HA (all but translating Dexippus’ text in the preceding lines) defines the relationship between Goths and Scythians. The words hoc est pars Gothorum, if read as “namely their Gothic part,” would correspond to the pattern described above. Cf. also S7.1–3 Martin (= fr.22.1–3 Mecella, FGrHist F 20–21.1–3) = HA Max. et Balb. 16.3: sub his pugnatum est a Carpis contra Moesos. fuit et Scyt⟨h⟩ici belli principium. 86 For this hypothesis see most recently M. Hose, “Historiographie in der Krise: Herausforderungen und Lösungen der Geschichtsschreibung im dritten Jahrhundert n. Chr.,” in Empire in Crisis 35–49, at 45–46. GUNTHER MARTIN AND JANA GRUSKOVÁ 471 ————— Greek, Roman, and Byzantine Studies 62 (2022) 437–493 Greek nominative would probably be Γουθούρικος, analogous to other Gothic names ending in -ος, e.g. Θεοδώριχος, Ἀλάριχος. These parallels also suggest that the accent was probably on the third syllable from the end. The ending -ικος may be a variant of the common ending -ιχος of Gothic names, cf. the Latin parallels such as Theodoric(h)us or Alaric(h)us. The other Gothic names occurring in the Scythica Vindobonensia, Cniva (Κνίβας, gen. Κνίβα) and Ostrogotha (Ὀστρόγουθθος or Ὀστρογούθθης)87 show a different kind of termination. The name Γουθούρικος is not attested elsewhere. The closest resemblance, in the extant sources, seems to be to Guntheri- cus,88 who, according to Jordanes,89 together with Argaithus was the ductor of the Gothic army in a campaign late in the reign of Philip (the so-called Arab), but the historicity of these events is controversial.90 The presence of nu would change the etymology from Guth/Goth = “Goth” to Gunth = “battle,” but the difference need not be an obstacle to the identification: the letter nu may have been omitted by a scribal error in the course of transmission, especially since this is a proper name (cf. above on Γοῦθοι); the etymological difference would have been meaningless to a Greek scribe and the name Γοῦθοι could be still echoing in his mind from earlier in the sentence. Forms 87 For the latest advances in the recovery of this name on fol. 194r see the forthcoming critical edition of the Scythica Vindobonensia. 88 Cf. H. Reichert, “Gunderich (§1 Namenkundliches),” in Reallex. Germ.Alt. 13 (1999) 194–195. 89 Jord. Get. 91–92 (XVI; ed. Giunta/Grillone): his ergo addens (scil. Ostro- gotha) Gothos et Peucinos (ab insula Peucis …), Argaithum et Gunthericum, nobilis- simos suae gentis ductores praefecit. qui mox Danubium vadati et de secundo Moesiam populati, Marcianopolim eiusdem patriae urbem, famosam metropolim adgrediuntur, diuque obsessam, accepta pecunia ab his qui inerrant, reliquerunt. 90 The name Gunthericus, blended with Argaithus, has often been suspected to be behind the form Argunth in HA Gord. 31.1. However, in the HA Argunth is Scytharum rex, Argaithus and Gunthericus in Jordanes are ductores under a Gothic king. 472 FACING THE PLAGUE AND THE GOTHS ————— Greek, Roman, and Byzantine Studies 62 (2022) 437–493 with and without nu could be interchanged, as is clear from the occasional insertion of the letter in several papyri rendering Aurelian’s title Gothicus as Γουνθικός (e.g. P.Ups.frid 6.18 [TM 15681, July/Aug. 273]; P.Oxy. XL 2902.18 [TM 45215, June 272]). Not least, forms starting with Gunth- are not attested before Gundomadus (Amm. Marc. 14.10.1 on the year 354). That means the name in the Scythica Vindobonensia may be cor- rect, whereas Jordanes altered it to conform to the preferences of his day. For the moment, we retain the text as transmitted in the palimpsest. The Scythica Vindobonensia informs us that Gouthourikos led (ἡγουµένου) the Goths in this invasion and commanded (ἦρχε̣) the entire army (τῆς πάσης στρατιᾶ̣ς̣). What the text does not provide―in contrast to the other two Gothic leaders mentioned ―is an exact status for Gouthourikos; cf. ὁ ἄρχων used for Ostrogotha (fol. 194r [fr. IIa] 17)91 and ὁ βασιλεύς for Cniva (fol. 195rv [fr. Iab]). All the same, it can be safely assumed that Gouthourikos was an eminent member of the Gothic elite, temporarily elected and accepted as leader of the Gothic army for this specific campaign. 28 Of φυγ(ῆ̣ς̣) at the end of line 28, the traces of the first three letters are visible enough. No other letter seems to follow in this line and the next line (29) starts with a new word. In the images, the area above φυγ is quite noisy, but there are some faint traces that could be part of an abbreviation for -ης, similar to φυλακ(ῆς) at the end of line 4 of fol. 194r (fr. IIa). We do not learn from the text why Gouthourikos was in exile. Was a failure in battle or defeat in a power struggle the reason,92 or was there something else?93 The use of the article 91 As Herwig Wolfram argues, Ostrogotha of the Scythica Vindobonensia is to be distinguished from the Amal king of the same name, Get. 91–92 (XVI): “Ostrogotha―ansischer Amaler oder glückloser Feigling,” in Empire in Crisis 17–34, at 30. 92 We now have the case of Ostrogotha (Scyth.Vind. fol. 194r [fr. IIa] 17ff.), GUNTHER MARTIN AND JANA GRUSKOVÁ 473 ————— Greek, Roman, and Byzantine Studies 62 (2022) 437–493 τῆς before φυγῆ̣ς̣ could (but need not necessarily) imply that there has been a reference to this exile earlier in the narrative. If the new invasion was the reason why Gouthourikos had been recalled from his exile, it could explain why Dexippus is point- ing to it. The text of the HA in the corresponding passage, Gall. 5.6, is corrupt. The authoritative codex P (Pal. lat. 899) offers the following wording (fol. 156r): gothori … a quo dictum est superius, Gothis inditum est.94 The word gothori is followed by a lacuna of approximately 17 letters, in which a later hand inserted non- sensical dodius. One might be tempted to ask whether the new evidence provides any clues for a reconstruction.95 ___ who seems concerned about the fact that he is being accused of failure in directing military operations, see Wolfram, in Empire in Crisis 17–34. It may be worth noting that “Argunth” is engaged in warfare against neighbouring tribes on the side of the Romans (possibly as an exile?), cf. HA Gord. 31.1 (n.90 above). 93 On exile among Germanic tribes see e.g. H. Holzhauer, “Verban- nung,” in Reallex.Germ.Alt. 32 (2006) 139–142. The issue will be discussed thoroughly by Herwig Wolfram in GLO 41–42 (in preparation). He argues: “Kaum ist anzunehmen, dass Guntherich wegen einer unglücklichen Niederlage verbannt wurde. Die Uraias-Geschichte lehrt, dass Unglück so stark ist, dass es in der Familie vererbt werden kann. Einem solchen ‘Unglücksmann’ wird nach der Verbannung kaum ein Heer anvertraut.” 94 Cf. Hohl: Gothoru⟨m pars⟩ . . . . . . . . . a quo dictum est superius, Gothis inditum est; Desbordes/Ratti: Gothi uel Getae, quod ⟨nomen, ut⟩ dictum est superius, Gothis inditum est. The newly discovered manuscript E (Erlangen Univ.-Bibliothek 647), which appears to be the only witness independent of P (see J. Stover, “New Light on the Historia Augusta,” JRS 110 [2020] 167–198), has an even larger lacuna in this place and cannot be used to fill the gap. 95 Franz Eyssenhardt (1864) suggested correcting the text (“indicante Sal- masio”) to Gothi sive Scythae quod nomen, ut dictum est superius, Gothis inditum est. Considering the new evidence provided by the Vienna palimpsest, one could offer the following reconstruction (pace legis Youtie): Gothori⟨co duce Scythae quod nomen,⟩ de (de ed. pr.: a P) quo dictum est superius, Gothis inditum est. Cf. also some lines later in the text, in 6.1 pugnatum est in Ach⟨a⟩ia Mariano duce contra eosdem Gothos. This conjecture, however, runs into a possible conflict 474 FACING THE PLAGUE AND THE GOTHS ————— Greek, Roman, and Byzantine Studies 62 (2022) 437–493 29–30 (+192v 1–2) ναυσὶ πορευθέντες ὑπὲρ̣ τοῦ | Ἴστρου τὸν πόρον εἰσέβαλλον εἴς τε Θρά⟨ι⟩||κην καὶ Μακεδονίαν καὶ ἐληίζοντο τὴν αὐτό|θι̣ γ̣ῆν σύµπασαν (“[the Goths] took the way across the Istros in ships, invaded Thrace and Macedonia and ravaged all the land there”): ναυσὶ πορευθέντες ὑπὲρ̣ τοῦ | Ἴστρου τὸν πόρον: in the transla- tion we interpret τὸν πόρον as accusative of space to πορευθέντες, with the specification ὑπὲρ̣ τοῦ Ἴστρου (LSJ s.v. πορεύω II.2. “c.acc.”; see also Montanari, GE s.v. 2.a). In an alternative interpretation, ὑπὲρ̣ would take the accusative τὸν πόρον, which would then denote one specific (well-known) crossing point of the Istros;96 the Goths would have entered the Istros from the Black Sea and sailed upstream beyond (ὑπὲρ̣) this crossing point ―presumably to avoid the Roman garrisons stationed there ―to land safely in Roman territory on their way to Thrace. The author obviously considered it worthwhile to mention that the Goths used ships to reach Roman territory. Never- theless, since the information is given without context, we cannot be sure about its function in the narrative. The infor- mation on ships may be retrospective, signalling an innovation in comparison with earlier invasions (e.g. if the Goths pre- viously crossed the Istros while it was frozen). Alternatively, the author may have wanted to introduce a skill of the Goths that was to play a more significant role later in the narrative―that means he was looking ahead to a period when ships became relevant (e.g. seaborne invasions of Asia Minor: HA Gall. 12.6 Scythae navibus factis Heracleam pervenerunt etc.). For further con- ___ with the identification of Dexippus’ Gouthourikos with Jordanes’ Gun- thericus: If the identification with Argunth in HA Gord. 31.1 (cf. n.90) is correct, the author of the HA would have used the person’s name once as “goth” and once as “gunth,” which seems improbable or would require a rather unusual combination of scribal mistakes. Bruno Bleckmann points out to us that the later scribe may have had in mind a text such as “Gothorido d(i)uce Scythae” with Gothorid being a variant of Gothorichus. 96 We owe this interpretation to Bruno Bleckmann and Nigel Wilson. GUNTHER MARTIN AND JANA GRUSKOVÁ 475 ————— Greek, Roman, and Byzantine Studies 62 (2022) 437–493 siderations see on Date below. εἰσέβαλλον εἴς τε Θρά⟨ι⟩||κην καὶ Μακεδονίαν: the conjecture to supply Θρᾴ]||κην, which we proposed in 2014,97 is shown to be correct. The parallel with HA Gall. 5.6 occupatis T⟨h⟩raciis Mace- doniam vastaverunt, Thessalonicam obsederunt is corroborated. The narrative moves at fast pace: the Gothic military actions, which must have taken weeks, are summed up in only 14 lines (fols. 192r [fr. IIIa] 29 to 192v [fr. IIIb] 12). Dexippus no more than mentions that the Goths invaded and plundered Thrace. As for the invasion into Macedonia, only six lines (fol. 192v [fr. IIIb] 2–7) are taken up by the unsuccessful siege of Thessaloni- ca, capital of the province.98 This brevity contrasts noticeably with the ekphrasis in which the stratagem and the nocturnal attack on the Thracian city Philippopolis is described in fr. Iab (fol. 195rv).99 The author concludes the section with the announcement of the Goths’ intention to advance on Athens and Achaia (fol. 192v [fr. IIIb] 7–12).100 The brevity of the section markedly differs from the lengthy and more analytical narrative of the de- fensive measures taken by the Greeks: the gathering of an army at Thermopylae,101 the election of the leaders, and Marianus’ speech. 97 Martin and Grusková, WS 127 (2014) 108. 98 For an overview of unsuccessful sieges of Thessalonica at that time see the commentary on Eusebius KFHist A6 fr.1 (=FGrHist 101 F 1) in B. Bleck- mann and J. Gross, Historiker der Reichskrise des 3. Jahrhunderts I (Paderborn 2016) 128–129. 99 For more on this and possible explanations see G. Martin in GLO 41– 42 (in preparation). 100 See Appendix I. The province of Achaia is referred to as ἡ εἴσω Πυλῶν Ἑλλάς, see fol. 192v (fr. IIIb) 23–24. 101 Cf. fol. 192v (fr. IIIb) 13ff.: ἐπεὶ δὲ̣ ἐ̣ς̣ τοὺς Ἕλληνας ἐξηγγέλθη ἡ ἔφοδος τ̣ῶν | Σκυθῶ̣ν̣ etc. (“And when the approach of the Scythians was reported to the Greeks” etc.); see Appendix I. 476 FACING THE PLAGUE AND THE GOTHS ————— Greek, Roman, and Byzantine Studies 62 (2022) 437–493 Inter- and intratextual relationships The new fragment bears an obvious relationship on a literary level to three other historiographical texts: two passages of Thucydides on the plague, HA Gall. 5–6, and the prooemium of Dexippus’ Skythika. The relationship of our fragment to each of them is different, as (a) serves Dexippus as a model, (b) fol- lows our text very closely, and (c) is a programmatic statement within the same work, so light is shed in different directions and on different aspects. a) Scyth. Vind. fr. IIIa (fol. 192r) 13–25 and Thucydides The fragments of the Scythica Vindobonensia have so far provided many examples of imitation of Thucydides in vocabulary, phrasing, syntax, and style.102 What had been absent is the sustained evocation of a situation or episode from the Thucydidean History, a type of imitation often identified in Dexippus’ ‘old’ fragments:103 for example, Decius’ letter to the defenders of Philippopolis recalls Pericles’ last speech, in which he tries to convince the Athenians to stay inside the walls and not to engage with the Spartans who are ravaging Attica―just as Decius tries to deter the Thracians, who are eager to take on the barbarian attackers in an open battle, from doing so.104 The new passage of the Scythica Vindobonensia now exhibits this trait of evoking specific episodes of Thucydides’ work. When 102 Cf. Martin and Grusková, WS 127 (2014) 115–116, and GRBS 54 (2014) 741–742; Mallan and Davenport, JRS 105 (2015) 207–209; A. Papathomas, “Dexippos und Thukydides,” in Empire in Crisis 135–146; C. M. Lucarini, “Textkritisches und Sprachliches zu Dexipp und zum Prosarhythmus der griechischen Historiker der Kaiserzeit,” in Empire in Crisis 73–94. 103 For one possible identification see H. Bannert, “Hoffen und Scheitern bei Thukydides und Dexippos,” in Empire in Crisis 53–62, at 59. 104 The same fragment borrows from Brasidas’ description of Scythians(!) (Thuc. 4.126.5). See F. J. Stein, Dexippus et Herodianus rerum scriptores quatenus Thucydidem secuti sint (Bonn 1957) 56–57; Martin, Dexipp von Athen 218–221. GUNTHER MARTIN AND JANA GRUSKOVÁ 477 ————— Greek, Roman, and Byzantine Studies 62 (2022) 437–493 Dexippus describes the plague, he seems to reuse material from two passages in which Thucydides reports similar scenarios. The underlined phrases mark verbal correspondences: Thuc. 2.47.3–4 (ed. Alberti): ἡ νόσος πρῶτον ἤρξατο γενέσθαι τοῖς Ἀθηναίοις, λεγόµενον µὲν καὶ πρότερον πολλαχόσε ἐγκατασκῆψαι καὶ περὶ Λῆµνον καὶ ἐν ἄλλοις χωρίοις, οὐ µέντοι τοσοῦτός γε λοιµὸς οὐδὲ φθορὰ οὕτως ἀνθρώπων οὐδαµοῦ ἐµνηµονεύετο γενέσθαι. οὔτε γὰρ ἰατροὶ ἤρκουν τὸ πρῶτον θεραπεύοντες ἀγνοίᾳ, ἀλλ’ αὐτοὶ µάλιστα ἔθνῃσκον ὅσῳ καὶ µάλιστα προσῇσαν, οὔτε ἄλλη ἀνθρω- πεία τέχνη οὐδεµία· ὅσα τε πρὸς ἱεροῖς ἱκέτευσαν ἢ µαντείοις καὶ τοῖς τοιούτοις ἐχρήσαντο, πάντα ἀνωφελῆ ἦν, τελευτῶντές τε αὐτῶν ἀπέστησαν ὑπὸ τοῦ κακοῦ νικώµενοι. Thuc. 3.87.1–4: Τοῦ δ’ ἐπιγιγνοµένου χειµῶνος ἡ νόσος τὸ δεύτερον ἐπέπεσε τοῖς Ἀθηναίοις, ἐκλιποῦσα µὲν οὐδένα χρόνον τὸ παντά- πασιν, ἐγένετο δέ τις ὅµως διοκωχή. παρέµεινε δὲ τὸ µὲν ὕστερον οὐκ ἔλασσον ἐνιαυτοῦ, τὸ δὲ πρότερον καὶ δύο ἔτη, ὥστε Ἀθηναίους γε µὴ εἶναι ὅ τι µᾶλλον τούτου ἐπίεσε καὶ ἐκάκωσε τὴν δύναµιν· τετρακοσίων γὰρ ὁπλιτῶν καὶ τετρακισχιλίων οὐκ ἐλάσσους ἀπέθανον ἐκ τῶν τάξεων καὶ τριακοσίων ἱππέων, τοῦ δὲ ἄλλου ὄχλου ἀνεξεύρετος ἀριθµός. ἐγένοντο δὲ καὶ οἱ πολλοὶ σεισµοὶ τότε τῆς γῆς… The highlighted vocabulary is partly so specific (ἐγκατασκῆψαι, διοκωχή)105 as to leave no doubt that Dexippus did not just use words and phrases that are generally applicable to the situa- tion, but reused for his scenario material from Thucydides’ report: for instance, both compare the present to former waves of the plague; the juxtaposition of sanctuaries and oracle(s) is another obvious borrowing. The passage in Thucydides’ Book 2 stands out as the kind of ‘purple passage’ that Dexippus uses elsewhere (like e.g. Pericles’ speeches), and he might presuppose familiarity with them in many of his readers. If we believe Lucian, the digression on the plague was imitated habitually by the emulating dilettanti of his 105 Between Thucydides and Dexippus, the word διοκωχή/διακωχή is found only in Cassius Dio and there only in the military sense (“armistice”), as has been mentioned in the commentary on lines 20–21 (with n.62). 478 FACING THE PLAGUE AND THE GOTHS ————— Greek, Roman, and Byzantine Studies 62 (2022) 437–493 day (about the Antonine plague: Hist.Conscr. 15). After Dexip- pus, Procopius gives his version concerning the Justinianic plague (Pers. 2.22–33, including the form ἐπέσκηψε in the same sense as ἐγκατασκῆψαι), and the Byzantine emperor and historian John Cantacuzenus reports a ‘Thucydidean’ plague for the fourteenth century.106 Dexippus in the new passage (fol. 192r [fr. IIIa] 17–25), which is likely to describe the climax of the disease, does not undertake to give a full description of the symptoms in the style of Thucydides, as his colleagues do. He may be indicating as much with the word “indescribable” (ἀδιήγητος)―unless this is mere hyperbole and he did mention the symptoms in the lost part of his work. However that may have been, the density and the unusual- ness of the material borrowed make it likely that Dexippus strove to produce an intertextual evocation of Thucydides’ account of the plague. One may wonder whether he aimed to achieve an effect beyond the display of learnedness and the insertion of himself into the tradition of Thucydidean emula- tion (as well as the insertion of his own time into the classical tradition of Athens). But it is clear that he is using the same technique of reminiscences as in other passages of the Skythika.107 b) Scyth. Vind. fr. III (fols. 192+193) and HA Gall. 5.2–6.1 There has been agreement that fr. III of the Scythica Vindo- bonensia is connected with the report in HA Gall. 5.2–6.1 (for the 106 III 49–53 Bonn. Cf. also Theon Progymn. (II 68.7–8 Spengel). See H. Hunger, “Thukydides bei Johannes Kantakuzenos. Beobachtungen zur Mimesis,” JÖB 25 (1976) 181–193; T. S. Miller, “The Plague in John VI Cantacuzenus and Thucydides,” GRBS 17 (1976) 385–395; M. Meier, “Beobachtungen zu den sogenannten Pestschilderungen bei Thukydides II 47–54 und bei Prokop, Bell. Pers. II 22–23,” Tyche 14 (1999) 177–210; Av. Cameron, Procopius and the Sixth Century (London 1985) 40–43. 107 This feature therefore further supports Dexippus’ authorship, if sup- port is still required. GUNTHER MARTIN AND JANA GRUSKOVÁ 479 ————— Greek, Roman, and Byzantine Studies 62 (2022) 437–493 text see Appendix II ).108 Through the new lines, the connection becomes all the more obvious. Here Dexippus provides the model that a later historian follows. The similarities are so sub- stantial as to rule out coincidence: the HA in part even pro- duces a near-verbal translation of the Greek. There are also differences: on the one hand, the HA omits several details, such as the sacrifices to Hera and Poseidon (fr. IIIa), the comparison between earlier stages of the plague and the present wave (fr. IIIa), the defensive measures at Ther- mopylae (fr. IIIb), the commanders Philostratus and Dexippus (fr. IIIc). That such pieces of information are missing may be explained by the free use of the source material on the part of the author of the HA. On the other hand, the HA mentions floodings of cities, crevices, and, most importantly, a summary of political events which is described as the raging of fortuna. At least part of these may have been dealt with in the first, still un- revealed part of fol. 192r or on the (preceding) now lost pages. At first glance, it may be tempting to assume that fr. III (fols. 192+193) of the Scythica Vindobonensia, and so Dexippus’ Skythika, is the direct source of HA Gall. 5.2–6.1. Since the HA typically appears to draw on one principal source for its ac- count of any period, that would mean by extension that the Skythika would be the source for the HA account from 238, the year in which Herodian’s history ended.109 There is, however, cause for circumspection. The HA narrative of the third quar- 108 On this connection see among others Grusková and Martin, in Das dritte Jahrhundert 267–281, esp. 269–270; Jones (n.14 above); Mallan and Davenport, JRS 105 (2015) 203–226; Piso, GFA 18 (2015) 205ff.; G. Zecchini, “Il nuovo Dexippo e l’Historia Augusta,” in B. Bleckmann et al. (eds.), Historiae Augustae Colloquium Dusseldorpiense (Bari 2017) 189–196, esp. 192–196. 109 T. D. Barnes, The Sources of the Historia Augusta (Brussels 1978) 109–112; Paschoud, in Historiae Augustae Colloquium Parisinum 217–269; D. Rohrbacher, “The Sources of the Historia Augusta Reexamined,” Histos 7 (2013) 146–180, at 165–166. 480 FACING THE PLAGUE AND THE GOTHS ————— Greek, Roman, and Byzantine Studies 62 (2022) 437–493 ter of the third century does not by any means have the in- vasions of the Germanic tribes as its focus, even if they occupy an important place. The material the HA includes is the barest outline of events as one might find in a chronicle. The frequent mention of consular dates in this part of the HA points to the use of a chronographical work as source. By contrast, Dexip- pus’ Skythika, being a monograph thematically focussed on wars against invading “Scythian” tribes (ending in Aurelian’s reign, after 271),110 surely did not provide extensive coverage of in- ternal Roman affairs, as the HA does.111 Hence, if Dexippus provided the principal source for the HA, it was―as has been repeatedly assumed112―not his Skythika, but his Chronika, a chronographic work of broad thematic scope covering the period from mythic times down to the reign of Claudius II (268–270). One may therefore infer that in those places where the HA is particularly close to the Scythica Vindobonensia, Dexip- pus’ Skythika was very similar to his Chronika.113 c) Scyth. Vind. fr. IIIa (fol. 192r) 13–25 and the prooemium of Dexippus’ Skythika As has been pointed out above, our passage relates back to the prooemium of the Skythika. We provide the relevant passage together with our own translation (Vat. gr. 73, fol. 54r [p. 107] 10–15):114 110 That the Skythika started in 238 is at least open to debate. Cf. G. Martin, “Die Struktur von Dexipps Skythika und die Historia Augusta,” in Historiae Augustae Colloquium Dusseldorpiense 97–114. 111 Cf. Martin, Dexipp von Athen 161–163; Mecella, Dexippo di Atene 96–112. 112 E.g. Barnes, Sources 109–112; Paschoud, in Historiae Augustae Colloquium Parisinum 217–269; B. Bleckmann, “Zu den Quellen der Vita Gallieni duo,” in Historiae Augustae Colloquium Maceratense 75–105; Mecella, Dexippo di Atene 29–34; Mallan and Davenport, JRS 105 (2015) 216, 221; Jones, in Empire in Crisis 162; Gengler, in Empire in Crisis 231. 113 For more see G. Martin in GLO 41–42 (in preparation). 114 This text was first published by A. Németh, in Empire in Crisis 125–128. For further observations see G. Martin (in preparation). GUNTHER MARTIN AND JANA GRUSKOVÁ 481 ————— Greek, Roman, and Byzantine Studies 62 (2022) 437–493 πρὸς δὲ ταῖς κατὰ πόλεµον συµ̣φ̣ο̣ρ̣|αῖς καὶ παθήµατα εἴρηταί µοι, ὅσ̣α συνηνέχθη γενέσθαι ἐν παν|τὶ τῶι πολέµωι τούτωι, ὧν ⁎⁎⁎⁎⁎⁎115 εὑρίσκω, ὅσα γε ἐν νόσωι τ̣ῆι λοι|µώδει καὶ σεισ̣µοῖς καὶ θαλάττης µεταβολῆι (ἣ δὴ κατὰ τὰς ἀµπώ̣|τεις ἠπειρώθη ἐνιαχῆι καὶ αὖθις κατὰ ἀθρόαν κύµατος ἐπιχώ|ρησιν προσβαλοῦσα ἔξω τοῦ τεταγµένου γῆν ἐπέσχε). In addition to the disasters caused by war, I have also spoken about sufferings, as far as they happened to occur during this entire war, sufferings of which I have found ⁎⁎⁎⁎⁎⁎, i.e. those in the plague, in earthquakes and changes of the sea (which in some places became land because of ebb tides, and then again with a complete advancement of the waves hit and occupied the land beyond its normal limits). At the start of his work, Dexippus announces in a program- matic statement that he will deal not only with the military history of the war against the “Scythians” and its concomitant disasters: he will also write about sufferings (παθήµατα) in a plague, earthquakes, and floods.116 Whilst in the rest of the fragments from the Skythika these phenomena are nowhere touched on, the new passage of the Vienna palimpsest men- tions two of the three παθήµατα of the prooemium (cf. HA Gall. 5.2–4): the plague is dealt with explicitly, while an earthquake can be confidently inferred from the reference to “Poseidon, Mover of the Earth.” Date The new information invites a discussion of the impact it has on the dating of the events mentioned. On the basis of fr. IIIb– IIIc (fols. 192v–193r) of the Scythica Vindobonensia, two dates are being considered in current scholarship: on the one hand the 115 The remnants of approximately six letters are too fragmentary to pro- pose a decipherment. 116 This itself imitates Thucydides (1.23.3): see the commentary on lines 17–18. 482 FACING THE PLAGUE AND THE GOTHS ————— Greek, Roman, and Byzantine Studies 62 (2022) 437–493 Athenian years 253/4117 or 254/5,118 on the other hand the early 260s.119 The former position rests on the information given in Syncellus (466.1–7), which places an invasion by the Scythians into Thrace, their unsuccessful attack on Thessa- lonica, and the fortification of Thermopylae early in the reign of Valerian and Gallienus;120 the latter is based on the consular date Gallieno et Faustiano consulibus in HA Gall. 5.2, i.e. 262. Both accounts agree in many substantial details with the Scythica Vindobonensia and therefore seem to refer to the same invasion, despite their contradictory dating. By offering some first thoughts we aim to encourage discussion of possible clues the new text may contain for the dating of the events described.121 1) As has been mentioned, the description of the plague 117 Orally this date had been proposed by Bruno Bleckmann in June 2014 before Christopher P. Jones independently argued for it (see n.14 above). For additional arguments in favour of 253/4 see e.g. C. M. Lucarini, “Zum neuen Dexipp,” ZPE 197 (2016) 42–45, at 45; Zecchini, in Historiae Augustae Colloquium Dusseldorpiense 189–190, D. Boteva, “Some Considerations Re- lated to the Scythica Vindobonensia,” in Empire in Crisis 195–212. 118 O. Gengler, “Eine neue Datierung des Goteneinfalls gegen Griechen- land unter Valerianus und Gallienus,” in Empire in Crisis 219–234, argues for 254/5, among other reasons, because in the following year Philostratus was archon and epinikia were organized in Athens. 119 E.g. Mallan and Davenport, JRS 105 (2015) 203, 215–222; Piso, GFA 18 (2015) 205–206, and “Das verhängnisvolle Jahr 262 und die amissio Daciae,” in L. Vagalinski et al. (eds.), Proceedings of the First International Roman and Late Antique Thrace Conference (Sofia 2018) 427–440, at 427–428; W. Eck, “Marianus, vice agens proconsulis Achaiae, im Dexippus Vindobonensis,” ZPE 208 (2018) 248–250. 120 Probably backed up by Zos. 1.29.2–3 and Zonar. 12.23 (III 139.26– 140.1 Dindorf), which derive from the same source. See B. Bleckmann, Die Reichskrise des III. Jahrhunderts in der spätantiken und byzantinischen Geschichts- schreibung: Untersuchungen zu den nachdionischen Quellen der Chronik des Johannes Zonaras (Munich 1992) 180–189. 121 For a very good overview of the current discussion on the dating of these events (including further bibliography) see Gengler, in Empire in Crisis 219–234. GUNTHER MARTIN AND JANA GRUSKOVÁ 483 ————— Greek, Roman, and Byzantine Studies 62 (2022) 437–493 appears to indicate that Dexippus is describing a major on- slaught or even the climax of the epidemic in Rome (and Greece) (see the commentary on lines 17–25). He refers to earlier waves and short intermissions between them. Two major difficulties arise in the attempt to use this piece of in- formation for dating. First, the uncertainty regarding the first outbreak of the ‘Plague of Cyprianʼ: while Kyle Harper assumes that the first wave hit Egypt in 249,122 recent con- tributions by Sabine Huebner and Christopher Jones present arguments that point to the reign of Trebonianus Gallus and Volusianus (251–253) as the time of the first wave.123 Second, while the number and length of the waves cannot be estab- lished,124 the date in the early 260s would provide more time for previous ones. Furthermore, it would chime with existing evidence for a strong wave around 260.125 2) Gouthourikos must have been an important figure among the Gothic leaders since he commanded the entire invading army (see the commentary on lines 27–29). The name, however, is new to us, and he cannot be confidently associated with any known character. Even if one supposes that the tentative identification with Jordanes’ Gunthericus, hinted at in the commentary, is correct, that does not help to exclude either of the above-mentioned dates: Gunthericus is said to have been one of the Gothic leaders in ca. 248, but the historicity of these events is controversial. If we accept Jordanes’ account and also identify Gunthericus with Gouthourikos, an absence of about 122 Harper, JRA 28 (2015) 227. 123 Huebner, JRA 34 (2021) 156–158; Jones, in Empire in Crisis 161. 124 If, as Dexippus indicates, there had been several διακωχαί before the one preceding the current wave, this must be at least the fourth wave. The intermissions themselves are characterised as “very brief.” 125 Cf. e.g. Zos. 1.37.3: λοιµὸς ἐπιβρίσας ταῖς πόλεσιν, οἷος οὔπω πρότερον ἐν παντὶ τῷ χρόνῳ συνέβη; Aur. Vict. Caes. 33.5: Simulque Romam pestilentia grassabatur, quae saepe curis gravioribus atque animi desperatione oritur. 484 FACING THE PLAGUE AND THE GOTHS ————— Greek, Roman, and Byzantine Studies 62 (2022) 437–493 five years may appear more likely than a recall after more than ten years. Any reliable conclusion, however, is precluded by the multiple uncertainties. 3) The mention of the fact that the Goths used ships for crossing the Istros suggests that this detail was noteworthy (see the commentary on lines 29–30). Yet, as no further information is given in fr. III, the function of this remark remains open. It appears that only incrementally did the Goths acquire the skill to use ships for their attacks and that only in the 260s were they fully able to do so.126 The brevity and vagueness of the refer- ence and the possible ambiguity in the wording (see the com- mentary) give cause for caution when trying to use the new information to pin down a date: if the lines about the Istros mean that the Goths crossed the river from the northern to the southern bank, such an operation would not require the same level of seamanship as sailing the Black and the Ionian Seas (as the “Scythians” did in the 260s: cf. e.g. HA Gall. 12.6 Scythae navibus factis Heracleam pervenerunt etc.). If, however, Dexippus intended to state that the Goths entered the Istros from the Black Sea and sailed upstream, this could still have taken place in the 250s: for, even then did the Goths sail to Asia, albeit not on their own but with the help of seafaring tribes.127 A con- clusion about the Goths’ navigational skills―and consequently about the date of the narrative―cannot therefore be drawn from our passage. 4) Finally, neither for the earthquake (see the commentary on lines 13–14 and 14–17) nor for the following consultation of the Sibylline Books do other historians provide independent 126 See e.g. Piso, GFA 18 (2015) 201–202. 127 As Zosimus reports, the invading tribes of the Black Sea region were able to make use of ships already in 253/4 (1.28.1, 1.31.1–2), but at that time they relied on the help provided by others. For a detailed analysis see A. Schwarcz, “Gotische und herulische maritime Einfälle in das Imperium Romanum,” in Empire in Crisis 389–401, at 390. GUNTHER MARTIN AND JANA GRUSKOVÁ 485 ————— Greek, Roman, and Byzantine Studies 62 (2022) 437–493 evidence that would allow a decision between the two dates proposed: They are otherwise attested only in HA Gall. 5.4–6 which, as seen, derives from Dexippus.128 Our discussion ends on an aporetic note. We have offered these preliminary thoughts to encourage further debate on the matter and look forward to new suggestions and assessments of clues for the dating of the historical events described in fr. III. APPENDIX I Fol. 192v (fr. IIIb) 2–16 (the continuation of the section under Edition):129 καὶ ἐπὶ τούτω̣⟨ι⟩ τῆ⟨ι⟩ Θεσσαλο|3νικ{⁎⁎}έων πόλει προσβαλόντες ἀθρό̣οι ἐπεί|4ραζον̣ α̣ὐ̣τῆς̣ τὴν̣ ἅλωσιν· ὡς δὲ ο̣ἵ̣ τε ἀπὸ |5 τοῦ τείχου̣ς̣ ε̣ὐ̣ρ̣ώ̣σ̣τ̣ω̣ς̣ ἠ̣µ̣ύ̣νοντ̣ο̣, πολυ̣χε̣ιρία⟨ι⟩ |6 τὰς τάξεις ἀµύν̣οντε̣ς̣, καὶ προ̣υ̣χώρ̣ει οὐ̣δὲν ἐς |7 ἐλπίδας, λύουσι τὴν πολι̣ο̣ρ̣κ̣ία̣ν. καὶ γνώ|8µη ⟨ἡ⟩ πλείστη̣ ἦ{εσ̣α}ν130 ἐπί τε̣ Ἀθήνας καὶ Ἀχ̣α̣ι̣ί|9αν̣ ὁ̣ρµ̣η̣θῆναι τῶ⟨ι⟩ στρατῶ⟨ι⟩, δόξη⟨ι⟩ τ̣ῶν ἐν τοῖς |10 Ἑλλη̣νικ̣ο̣ῖ̣ς ἱερ̣οῖς̣ χρ̣υ̣σ̣ῶν̣ καὶ ἀ̣ρ̣γυρῶν ἀνα|11θη̣µά̣τ̣ω̣ν̣ καὶ ὅσ̣α ποµ̣πε̣ῖ̣α̣· πλουσιώτατον |12 γὰ̣ρ τα̣ύτη⟨ι⟩ τὸ χωρίον εἶ̣ναι̣ ἐξεπυν̣θά̣νον̣τ̣ο̣. |13 ἐπεὶ δὲ̣ ἐ̣ς̣ τοὺς Ἕλληνας ἐξηγγέλθη ἡ ἔφοδος τ̣ῶν |14 Σκυθῶ̣ν̣, συ̣ν̣ή̣ε̣σ̣α̣ν̣ ἐ̣ς Πύλας καὶ κατὰ τα̣ὐ̣|15τόθι στεν̣ὰ̣ τῶν̣ π̣α̣ρόδων̣ ἐξείρ- γειν αὐτοὺς ὥρ|16µη̣ντο.… And then they attacked the city of Thessalonica with their full force and tried to take it. But when those on the walls fought them off valiantly, warding off the attacking divisions by their own large number, and they [i.e. the Scythians] made no progress in their hopes, they lifted the siege. And the army was most inclined to ad- 128 So far archaeological and epigraphic evidence does not seem to be datable with sufficient precision as to help dating these events. 129 This work-in-progress edition is based on the preliminary transcription published by Martin and Grusková, WS 127 (2014) 106; reprinted in Martin and Grusková, in Empire in Crisis 547–548. We will present the final edition of this passage in the critical edition (editio princeps) of the Scythica Vindobonensia. For further suggestions on the text see e.g. Mallan and Davenport, JRS 105 (2015) 205–206; Jones (n.14 above); Piso, GFA 18 (2015) 200–201, 207; Lucarini, ZPE 197 (2016) 42–45. 130 For this preliminary emendation of ηεσ̣αν (Cod.) see Martin and Grusková, WS 127 (2014) 109. 486 FACING THE PLAGUE AND THE GOTHS ————— Greek, Roman, and Byzantine Studies 62 (2022) 437–493 vance on Athens and Achaia, imagining the gold and silver dedica- tions in the Greek sanctuaries and all the processional objects. For they were hearing reports that the region there was very wealthy. When the Scythians’ advance was reported to the Greeks, they gathered at Thermopylae and set themselves to block them using the narrow passes there.… APPENDIX II HA Gall. 5.2–7, 6.1–2 (ed. Hohl with two adaptations):131 5 (2) Gallieno et Faustiano132 conss. inter tot bellicas clades etiam terrae motus gravissimus fuit et tenebrae per multos ⟨dies⟩, (3) auditum praeterea tonitruum terra mugiente, non Iove tonante, quo motu ipsae multae fabricae devoratae sunt cum habitatoribus, multi terrore emortui; quod quidem malum tristius in Asiae urbibus fuit. (4) mota est et Roma, mota Libya. hiatus terrae plurimis in locis fuerunt, cum aqua salsa in fossis appareret. maria etiam multas urbes occu- parunt. (5) pax igitur deum quaesita inspectis Sibyll⟨a⟩e libris factu⟨m⟩que Iovi Salutari, ut praeceptum fuerat, sacrificium. nam et pestilentia tanta extiterat vel Romae vel in Achaicis urbibus, ut uno die quinque milia hominum pari morbo perirent. (6) Saeviente fortuna, cum hinc terrae motus, inde hiatus soli, ex diver- sis partibus pestilentia orbem Romanum vastaret, capto Valeriano, Gallis parte maxima opsessis, cum bellum Odenatus inferret, cum Aureolus perurgeret . . . . cum ⟨A⟩emilianus Aegyptum occupasset, Gothoru⟨m pars⟩ . . . . . . . . . a quo dictum est superius, Gothis inditum est, occupatis T⟨h⟩raciis Macedoniam vastaverunt, Thessalonicam obsederunt, neque usquam quies mediocriter sal[u]tem ostentata est. (7) quae omnia contemptu, ut saepius diximus, Gallieni fiebant, hominis luxuriosissimi et, si esset securus, ad omne dedecus paratissimi. 6 (1) Pugnatum est in Ach⟨a⟩ia Mariano133 duce contra eosdem Gothos, unde victi per Ach⟨a⟩eos recesserunt. (2) Scythae autem, hoc est pars Gothorum, Asiam vastabant. Acknowledgments Since the discovery of the Scythica Vindobonensia in Cod. hist. gr. 73 (fols. 192r–195v) of the Austrian National Library in Vienna,134 the revelation, 131 We print everything in italics, whereas Hohl distinguishes between trans- mitted text and editorial departures from P. 132 Hohl prints Faus[t]iano. The correctness of the transmitted Faustiano has been demonstrated by Mallan and Davenport, JRS 105 (2015) 216. 133 Hohl prints Mar⟨c⟩iano. Thanks to the evidence of the Scythica Vindobonensia, this conjecture can be definitively rejected. 134 The fragments were discovered in 2007–2009 by Jana Grusková during a GUNTHER MARTIN AND JANA GRUSKOVÁ 487 ————— Greek, Roman, and Byzantine Studies 62 (2022) 437–493 decipherment, and edition of the new fragments have been the objective of two major research projects funded by the Austrian Science Fund (FWF). To render accessible the otherwise invisible writing of the eleventh-century manuscript, we cooperated with imaging teams specialized in digital recovery of palimpsests. The Austrian National Library, represented by Andreas Fingernagel, Katharina Kaska (since 2015), Ernst Gamillscheg (until 2015), and Christa Hofmann, generously made the manuscript accessible. During all our experiments, highest priority was given to ensuring the continued safety of the written historical artefact.135 (1) In the period from 2012 to 2015, in the project “Important textual witnesses in Vienna Greek palimpsests” (FWF P 24523-G19), which was led by Otto Kresten and hosted at the Austrian Academy of Sciences, Department of Byzantine Research (Institute for Medieval Research),136 multispectral imaging and image processing were performed in coopera- tion with Early Manuscripts Electronic Library and a team of image scientists and engineers assembled by Michael Phelps, director of EMEL, for this project.137 Based on the multispectral images that were captured in 2013 and processed in 2013–2015,138 we were able to decipher and preliminarily publish about 60% of the palimpsest. (2) Since 2015, we have been able to further pursue the decipherment with a view to a complete critical edition (editio princeps) of the Scythica ___ research project of the Austrian Academy of Sciences (Byzanzforschung), which focused on Greek palimpsests of the Austrian National Library. The project was led by Otto Kresten. By using technical means available at that time, primarily an ultraviolet lamp and old UV photographs, Grusková managed to decipher 15% of the text; see Grusková, Untersuchungen 50–53. 135 For more details see J. Grusková, G. Martin, and O. Kresten, “Scythica Vindobonensia: Geschichte und Ausblick,” AnzWien 153 (2018 [2019]) 69–92, and J. Grusková, G. Martin, O. Kresten, F. Mitthof, K. Kaska, Ch. Hofmann, W. Kreuzer, M. Phelps, K. Boydston, R. L. Easton, Jr., K. T. Knox, D. Kelbe, D. Kasotakis, W. A. Christens-Barry, D. Stewart, I. Rabin, O. Hahn, L. Glaser, J. Garrevoet, I. Shevchuk, S. Klumpp, D. Deckers, and J. Buck, “Insights into the Digital Recovery of the Scythica Vindobonensia,” in M. Cronier et al. (eds.), Le livre manuscrit grec: écritures, matériaux, histoire (Paris 2020) 945–967. 136 See Grusková et al., AnzWien 153 (2018) 75–78. 137 For further details and acknowledgments see Grusková et al., in Empire in Crisis 549–550, and Grusková et al., in Le livre manuscrit grec 946–947, 950–953. 138 See fig. 1–6 in Grusková et al., in Empire in Crisis 551–556. 488 FACING THE PLAGUE AND THE GOTHS ————— Greek, Roman, and Byzantine Studies 62 (2022) 437–493 Vindobonensia. In the period 2015 to 2020, our research was carried out in the project “Scythica Vindobonensia” (FWF P 28112-G25). It was conducted jointly at the Department of Ancient History, Papyrology and Epigraphy of the University of Vienna and the Department of Byzantine Research (IMAFO) of the Austrian Academy of Sciences, and led by Fritz Mitthof with Otto Kresten as national research partner. That project also aimed at a broader historiographical, historical, and archaeological contextualization of the new text.139 In the course of this project, a new round of multispectral imaging was conducted in 2016 in cooperation with EMEL’s team: Michael Phelps, Ken Boydston, Roger L. Easton, Jr., Keith T. Knox, David Kelbe, Damianos Kasotakis, Dale Stewart.140 In the same year, we performed first experiments with X-ray fluorescence (XRF) element mapping at the Bundesanstalt für Materialforschung und -prüfung in Berlin in coopera- tion with Ira Rabin and Oliver Hahn (BAM/CSMC) to see if this method could achieve further recovery.141 Finally, in 2017 we experi- mentally applied fast-scanning XRF element mapping to the palimpsest at the Deutsches Elektronen-Synchrotron DESY in Hamburg (Proposal I-20170505 EC) with Jana Grusková as the project leader and Leif Glaser, an expert in material analysis of artefacts and archaeometry physics, as the principal investigator. DESY (Hamburg, Germany), a member of the Helmholtz Association HGF, generously provided the experimental facilities. The research was carried out at PETRA III at the P06 Beamline Hard X-Ray Micro-Probe.142 The loan of the valuable manuscript from the Library and the expenses of its trip to DESY were covered by a grant from the Austrian Academy of Sciences (“Holzhausen-Legat”) awarded to Fritz Mitthof, on behalf of the project “Scythica Vindobonensia.” The two illegible pages, fols. 192r and 193v, 139 See Grusková et al., AnzWien 153 (2018 [2019]) 78–81. For further details see the homepage https://www.oeaw.ac.at/scythica-vindobonensia/ (last accessed 25 Nov. 2022). See also F. Mitthof et al., Empire in Crisis. 140 See Grusková et al., in Le livre manuscrit grec 947–948, 953–964. 141 See Grusková et al., in Le livre manuscrit grec 956–960. 142 On the P06 Beamline in general see Ch. G. Schroer, P. Boye, J. M. Feldkamp, P. Patommel, D. Samberg, A. Schropp, A. Schwab, S. Stephan, G. Falkenberg, G. Wellenreuther, and N. Reimers, “Hard X-ray Nanoprobe at Beamline P06 at PETRA III,” Nuclear Instruments & Methods in Physics Research A 616 (2010) 93–97. For further details see Grusková et al., in Le livre manuscrit grec 956–962. GUNTHER MARTIN AND JANA GRUSKOVÁ 489 ————— Greek, Roman, and Byzantine Studies 62 (2022) 437–493 and other problematic areas were imaged at DESY in December 2017. The state-of-the-art expertise in image science required for the processing of the large amount of data collected since 2013 (both MSI and XRF) was provided by Roger L. Easton, Jr., Keith T. Knox, and (until 2018) David Kelbe, the image scientists assembled by EMEL for our projects.143 Work on the present fol. 192r has been particularly painstaking and intensive, in particular since 2018. The recovery of the text often had to proceed in very small cropped sections, in constant exchange between Roger L. Easton, Jr., and Keith T. Knox on the one side and the authors of this article on the other.144 The image scientists spared no effort to test a wide range of different processing approaches to offer the best results that this difficult palimpsest could yield. They were always willing to respond to our feedback and perform further processing of the image data. The authors of this article would like to express their deepest gratitude to all participating scientists and institutions.145 November, 2022 Institut für Klassische Philologie Heinrich-Heine-Universität Düsseldorf gunther.martin@hhu.de Seminar für griech. und lat. Philologie Universität Zürich gunther.martin@uzh.ch Abteilung Byzanzforschung IMAFO Österreichische Akademie der Wissenschaften jana.gruskova@oeaw.ac.at Katedra klasickej a semitskej filológie FiF Univerzita Komenského v Bratislave jana.gruskova@uniba.sk 143 In 2013–2014, they were joined by William A. Christens-Barry. 144 See Grusková et al., in Le livre manuscrit grec 956–960, and a paper on image processing, which is being prepared by the authors of this article and the scientists Roger L. Easton, Jr., Keith T. Knox, Leif Glaser, Katharina Kaska, David Kelbe, and Ivan Shevchuk. 145 We would like to thank Fritz Mitthof, Otto Kresten, Herwig Wolfram, Bruno Bleckmann, Markus Stein, Herbert Bannert, and both the anonymous and the named―Nigel Wilson and Carlo M. Lucarini―referees of GRBS for their most valuable feedback on drafts of this article. 490 FACING THE PLAGUE AND THE GOTHS ————— Greek, Roman, and Byzantine Studies 62 (2022) 437–493 Figure 1: Vienna, ÖNB, Codex hist. gr. 73, fol. 192r. “Visual appearance image” by the Early Manuscripts Electronic Library. © ÖAW Project FWF P 24523-G19 GUNTHER MARTIN AND JANA GRUSKOVÁ 491 ————— Greek, Roman, and Byzantine Studies 62 (2022) 437–493 Figure 2: Vienna, ÖNB, Codex hist. gr. 73, fol. 192r. Spectral imaging by the Early Manuscripts Electronic Library. Processed image by Roger L. Easton, Jr. © ÖAW Project FWF P 24523-G19 492 FACING THE PLAGUE AND THE GOTHS ————— Greek, Roman, and Byzantine Studies 62 (2022) 437–493 Figure 3: Vienna, ÖNB, Codex hist. gr. 73, fol. 192r undertext: Scythica Vindobonensia, fr. IIIa 13–30. XRF element mapping at DESY by Leif Glaser (Project I-20170505 EC). Processed image by Keith T. Knox (EMEL). © ÖAW Project FWF P 28112-G25 GUNTHER MARTIN AND JANA GRUSKOVÁ 493 ————— Greek, Roman, and Byzantine Studies 62 (2022) 437–493 Figure 4: Vienna, ÖNB, Codex hist. gr. 73, fol. 192r undertext: Scythica Vindobonensia, fr. IIIa 13–30. XRF element mapping at DESY by Leif Glaser (Project I-20170505 EC). Processed image by Roger L. Easton, Jr. (EMEL). © ÖAW Project FWF P 28112-G25