Chronicles, Consuls, and Coins: Historiography and History in the Later Roman Empire (Variorum Collected Studies) [1 ed.] 9781409428206, 1409428206

Citation preview

Also in the Variorum Collected Studies Series:

LESLIE S.B. MACCOULL Documenting Christianity in Egypt, Sixth to Fourteenth Centuries

CLAIRE SOTINEL Church and Society in Late Antique Italy and Beyond

JOHN WORTLEY Studies on the Cult of Relics in Byzantium up to 1204

NEIL MCLYNN Christian Politics and Religious Culture in Late Antiquity

JAMES HOWARD-JOHNSTON East Rome, Sasanian Persia and the End of Antiquity Historiographical and Historical Studies

ROGER S. BAGNALL Hellenistic and Roman Egypt Sources and Approaches

J.H.W.G. LIEBESCHUETZ Decline and Change in Late Antiquity Religion, Barbarians and their Historiography

HENRY CHADWICK Studies on Ancient Christianity

MARK VESSEY Latin Christian Writers in Late Antiquity and their Texts

ROGER S. BAGNALL Later Roman Egypt: Society, Religion, Economy and Administration

YVES-MARIE DUVAL Histoire et historiographie en Occident aux IVe et Ve siècles

ROBERT A. MARKUS Sacred and Secular Studies on Augustine and Latin Christianity

LESLIE S.B. MACCOULL Coptic Perspectives on Late Antiquity

VARIORUM COLLECTED STUDIES SERIES

Chronicles, Consuls, and Coins: Historiography and History in the Later Roman Empire

R.W. Burgess

Chronicles, Consuls, and Coins: Historiography and History in the Later Roman Empire

First published 2011 by Ashgate Publishing Published 2016 by Routledge 2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN 711 Third Avenue, New York, NY 10017, USA Routledge is an imprint of the Taylor & Francis Group, an informa business This edition © 2011 by R.W. Burgess R.W. Burgess has asserted his moral right under the Copyright, Designs and Patents $FWWREHLGHQWL¿HGDVWKHDXWKRURIWKLVZRUN

All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. British Library Cataloguing in Publication Data Burgess, R. W. (Richard W.) Chronicles, consuls, and coins : historiography and history in the later Roman Empire. – (Variorum collected studies series) 1. Rome – History – Empire, 284–476 – Historiography. 2. Rome – History – Empire, 284–476 – Sources. I. Title II. Series 937'.08–dc22 ISBN 978–1–4094–2820–6 Library of Congress Control Number: 2011925098 ISBN 9781409428206 (hbk)

VARIORUM COLLECTED STUDIES SERIES CS984

CONTENTS Introduction

vii

Acknowledgements

ix

Abbreviations

xi

HISTORIOGRAPHY I

The dates and editions of Eusebius’ Chronici canones and Historia ecclesiastica

471–504

Journal of Theological Studies, n.s. 48, 1997

II

A chronological prolegomenon to reconstructing Eusebius’ Chronici canones: the evidence of Ps-Dionysius (the Zuqnin Chronicle)

29–38

Journal of the Canadian Society for Syriac Studies 6, 2006

III

Jerome explained: an introduction to his Chronicle and a guide to its use

1–32

Ancient History Bulletin 16, 2002

IV

Jerome and the Kaisergeschichte

349–369

Historia 44, 1995

V

On the date of the Kaisergeschichte

111–128

Classical Philology 90, 1995

VI

Principes cum tyrannis: two studies on the Kaisergeschichte and its tradition

491–500

Classical Quarterly 43, 1993

VII

A common source for Jerome, Eutropius, Festus, Ammianus, and the Epitome de Caesaribus between 358 and 378, along with further thoughts on the date and nature of the Kaisergeschichte Classical Philology 100, 2005

166–192

vi

VIII

CONTENTS

Eutropius v.c. magister memoriae?

76–81

Classical Philology 96, 2001

HISTORY IX

X

Ѡ͔͎͒Ӊ͋LO̯͎͍ҽ͇͍͐͑̓͋? The location and circumstances of Constantine’s death Journal of Theological Studies, n.s. 50, 1999 The summer of blood: the ‘great massacre’ of 337 and the promotion of the sons of Constantine

153–161

5–51

Dumbarton Oaks Papers 62, 2008

XI

The Passio S. Artemii, Philostorgius, and the dates of the invention and translations of the relics of Sts Andrew and Luke

5–36

Analecta Bollandiana 121, 2003

XII

The accession of Marcian in the light of Chalcedonian apologetic and monophysite polemic

47–68

Byzantinische Zeitschrift 86/87, 1993/1994

XIII

The third regnal year of Eparchius Avitus: a reply

335–345

Classical Philology 82, 1987

XIV

Quinquennial vota and the imperial consulship in the IRXUWKDQG¿IWKFHQWXULHV± ±SODWHV Numismatic Chronicle 148, 1988

XV

‘Non duo Antonini sed duo Augusti’: the consuls of 161 and the origins and traditions of the Latin consular fasti of the Roman Empire

259–290

Zeitschrift für Papyrologie und Epigraphik 132, 2000

Supplementary Notes Index

1–10 1–5

This volume contains xiv + 354 pages

INTRODUCTION The papers collected in this volume extend from my second published paper LQEHIRUH,KDG¿QLVKHGP\'3KLOWKHVLVWRP\PRVWUHFHQWSDSHUZKLFK DSSHDUHG LQ PLG LQ VSLWH RI LWV SXEOLFDWLRQ GDWH RI  2QO\ WZR RI them, papers XII and XIV, did not arise out of my main research interest, ZKLFKLVFKURQLFOHV7KH\ZHUHWKHUHVXOWRIP\¿UVWSODQIRUDGRFWRUDOWKHVLV WRSLFDKLVWRULFRQXPLVPDWLFDQDO\VLVRIWKHKLVWRU\RIWKH:HVWEHWZHHQ DQG6RPHRIWKDWRULJLQDOQXPLVPDWLFUHVHDUFKFDQEHVHHQSHHNLQJRXW LQQRISDSHU;,,,DVZHOO$QLQWHUHVWLQFRQVXOVDQGWKHLUDFFHSWDQFHDQG QRQDFFHSWDQFHDVDZLQGRZLQWR(DVW:HVWUHODWLRQVZDVDSDUWRIWKDWRULJLQDO WKHVLVSODQDQGWKDWHDUO\UHVHDUFKOHGWRP\¿UVWSXEOLVKHGSDSHUµ7KHQLQWK FRQVXOVKLS RI +RQRULXV$'  DQG ¶ Zeitschrift für Papyrologie und Epigraphik >@± DVZHOODVSDSHUV;,9DQGHYHQWXDOO\ ;9 7KDWRULJLQDOWKHVLVSODQHYHQWXDOO\JDYHZD\WRDFRPSOHWHO\GLIIHUHQWWRSLF WKHFKURQLFOHRI+\GDWLXVDQGQXPLVPDWLFVKDGWRWDNHDEDFNVHDWIRUDIHZ GHFDGHV,WLVIURPWKDWWKHVLVDQGWKHSXEOLFDWLRQRILWVVHFRQGYROXPHE\WKH &ODUHQGRQ3UHVVLQ WKDWSDSHU;,,,DQGDOOP\ODWHUZRUNKDVHYROYHGLQ RQHZD\RUDQRWKHUWKHVWXG\RI+\GDWLXV¶VRXUFHVIRUWKHWKHVLVKDGDOUHDG\ led to the Descriptio consulum ZKLFK ZDV WKHQ LQFOXGHG ZLWK P\  edition of Hydatius, and the Descriptio then led to other consulariaZKHQFH HYHQWXDOO\SDSHU;, DQGDWUDQVODWLRQRIDQGFRPPHQWDU\RQ+\GDWLXVVWDUWHG LQOHGWR-HURPHDQGHYHQWXDOO\SDSHU,,, ±LQWHQGHGDVDSUROHJRPHQRQ WRWKH+\GDWLXVYROXPH±ZKLFKOHGDVWXG\RI-HURPH¶VVRXUFHVZKLFKOHG WR(XVHELXVDQGWKH$QWLRFKHQHFRQWLQXDWLRQRI(XVHELXVSDSHU,DVZHOODV Studies in Eusebian and Post-Eusebian Chronography>6WXWWJDUW@ZKLFK WKHQOHGWRSDSHUV,,,;DQG; DQGWKHKaisergeschichteSDSHUV,9±9,,DQG LQGLUHFWO\9,,, 0DQ\RIWKHVHDUWLFOHVKDGDORQJJHQHVLVWKURXJKWKHZULWLQJ DQGSXEOLFDWLRQSURFHVVSDSHU;,,ZDVZULWWHQLQWKHODWHVDQGGHOD\HG E\ D EDFNORJ DW Byzantinische Zeitschrift SDSHU ,,, WKH µ'XPP\¶V JXLGH WR -HURPH¶DVLWZDVRULJLQDOO\FDOOHGDQGE\ZKLFKWLWOH,VWLOOUHIHUWRLWZDV EHJXQLQDQGZDVLQWKHHQGGHOD\HGE\WZRVHWVRIUHIHUHHVDQGHGLWRUV ZKRHDFKWRRND\HDUWRUHQGHUDYHUGLFWZKLFKZDVQHJDWLYHEHFDXVHRILWV OHQJWK DQGSDSHU;ZDVVWDUWHGLQGXULQJDVDEEDWLFDOLQ2[IRUGZKLOH, ZDVZRUNLQJRQSDSHUV,;DQG;9DQGZDVLQWKHHQGGHOD\HGE\SURGXFWLRQ GLI¿FXOWLHVDWDumbarton Oaks Papers.

viii

INTRODUCTION

The central focus of the papers collected in this volume is the sources for UHFRQVWUXFWLQJWKHKLVWRU\RIWKHWKLUGWR¿IWKFHQWXULHV AD7KH¿UVWVHFWLRQ µ+LVWRULRJUDSK\¶FRQFHQWUDWHVRQWKHHOXFLGDWLRQRIDVPDOOJURXSRIFKURQLFOHV and breviaria whose texts are fundamental for our reconstruction of the KLVWRU\RIWKHWKLUGDQGIRXUWKFHQWXULHVVRPHZHOONQRZQRWKHUVPXFKOHVV VR (XVHELXV RI &DHVDUHD -HURPH WKH ORVW Kaisergeschichte DQG (XWURSLXV ,QWKLVVHFWLRQWKHJRDOLQHDFKFDVHLVDVSHFL¿FDWWHPSWWRFRPHWRDEHWWHU XQGHUVWDQGLQJRIWKHVWUXFWXUHFRPSRVLWLRQGDWHRUDXWKRURIWKHVHKLVWRULFDO texts. 7KH VHFRQG VHFWLRQ µ+LVWRU\¶ SUHVHQWV D JURXS RI KLVWRULFDO VWXGLHV ranging in time from the death of Constantine in 337 to the vicennalia of $QDVWDVLXV LQ  ,Q WKHVH SDSHUV WKH NH\V WR WKH FRQFOXVLRQV RIIHUHG DULVH from a better understanding of the literary sources – particularly chronicles and consularia±DQXQGHUVWDQGLQJRIWKHHYROXWLRQRIKLVWRULFDODFFRXQWVRYHU WLPH RU WKH HPSOR\PHQW RI VRXUFHV WKDW DUH HLWKHU QHZ RU XQXVXDO LQ WKHVH particular contexts: consular fastiFRLQVSDS\ULDQGLWLQHUDULHV R.W. BURGESS Ciuitate Ottauiense V id. Ian. MMXI

ACKNOWLEDGEMENTS Grateful acknowledgement is made to the following persons, institutions, journals and publishers for their kind permission to reproduce the papers included in this volume: Oxford University Press (for Papers I and IX); the Canadian Society for Syriac Studies, University of Toronto (II); Ancient History Bulletin, Calgary (III); Franz Steiner Verlag, Stuttgart (IV); University of Chicago Press (V, VII, VIII, XIII); Cambridge University Press (VI); Dumbarton Oaks Papers, Washington DC (X); the Société des Bollandistes (Analecta Bollandiana), Brussels (XI); Walter de Gruyter GmbH & Co., Berlin (XII); the Royal Numismatic Society (Numismatic Chronicle) (XIV); and Dr Rudolf Habelt GmbH, Bonn (XV).

PUBLISHER’S NOTE The articles in this volume, as in all others in the Variorum Collected Studies Series, have not been given a new, continuous pagination. In order to avoid confusion, and to facilitate their use where these same studies have been referred to elsewhere, the original pagination has been maintained wherever possible. Each article has been given a Roman number in order of appearance, as listed in the Contents. This number is repeated on each page and is quoted in the index entries. Asterisks in the margins are to alert the reader to additional information supplied at the end of the volume in the Supplementary Notes.

ABBREVIATIONS AB ACOec AE/AEpigr AJAH AJPh ANSMN BASP BFLM BHAC BICS BJ BullBudé Byz ByzF BZ CB CCSL Chron. min. CIL CLRE

CP/CPh CQ CR CSCO CSHB CTh DOP EMC/CV

Analecta Bollandiana Acta Conciliorum Oecumenicorum. Leipzig and Berlin, 1914–. L’Année épigraphique American Journal of Ancient History American Journal of Philology American Numismatic Society Museum Notes Bulletin of the American Society of Papyrologists Bulletin de la Faculté des Lettres de Mulhouse Bonner Historia-Augusta-Colloquium Bulletin of the Institute of Classical Studies Bonner Jahrbücher Bulletin de l’Association Guillaume Budé Byzantion Byzantinische Forschungen Byzantinische Zeitschrift The Classical Bulletin Corpus Christianorum series Latina. Turnholt, 1953–. Chronica minora (= MGH: AA vols. 9, 11, and 13) Th. Mommsen, et al. Corpus Inscriptionum Latinarum. Berlin, 1869–. Roger S. Bagnall, Alan Cameron, Seth R. Schwartz, and Klaas A. Worp. Consuls of the Later Roman Empire. Atlanta, 1987. Classical Philology Classical Quarterly Classical Review Corpus scriptorum christianorum orientalium. Paris and Louvain, 1903–. Corpus scriptorum historiae Byzantinae. Bonn, 1828–1897. Th. Mommsen and P. M. Meyer. Theodosiani Libri XVI cum constitutionibus Sirmondianis. Berlin, 1905. Dumbarton Oaks Papers Echos du monde classique/Classical Views

xii

GCS

ABBREVIATIONS

Die griechischen christlichen Schriftsteller der ersten Jahrhunderte. Leipzig and Berlin, 1897–. GOTR Greek Orthodox Theological Review GRBS Greek, Roman and Byzantine Studies HE Historia ecclesiastica HCPh/HSCP Harvard Studies in Classical Philology HLL R. Herzog and P.L. Schmidt (eds). Handbuch der lateinischen Literatur der Antike. Munich, 1989–. HTR Harvard Theological Review ICUR Giovanni Battista de Rossi, Angelo Silvagni, and Antonio Ferrua. Inscriptiones Christianae urbis Romae septimo saeculo antiquiores. Nova series. Rome, 1922–. ILS Hermann Dessau (ed.). Inscriptiones Latinae Selectae. Berlin, 1892–1916 Inscr. Ital. Inscriptiones Italiae. Rome, 1931–. JEH Journal of Ecclesiastical History JHS Journal of Hellenic Studies JIAN Journal international d’archéologie numismatique JLA Journal of Late Antiquity JÖBG Jahrbuch der österreichischen byzantinischen Gesellschaft JRS Journal of Roman Studies JTS/JThS Journal of Theological Studies LRBC R.A.G. Carson, P.V. Hill, and J.P.C. Kent. Late Roman Bronze Coinage. London, 1978. Henry George Liddell, Robert Scott, and Henry Stuart Jones LSJ 9 (eds.). A Greek English Lexicon.9 Oxford, 1940. MGH: AA Monumenta Germaniae Historica: Auctores Antiquissimi. Berlin, 1877–1919. MGH: SSRLang Monumenta Germaniae historica: Scriptores rerum Langobardicarum et Italicarum saec. VI–IX. Berlin, 1878. MGH: SSRM Monumenta Germaniae historica: Scriptores rerum Merovingicarum. Berlin, 1885–1951. NC Numismatic Chronicle NNÅ Nordisk Numismatisk Årsskrift NNM Numismatic Notes and Monographs NZ Numismatische Zeitschrift PBA Proceedings of the British Academy PG Patrologia graeca PIR Elimar Klebs, et al. Prosopographia imperii romani saec I. II. III.1 and 2 Berlin, 1897–. PL Patrologia latina

ABBREVIATIONS

PLRE PO RBN RE REA RendLinc  RFIC RhM/RM RIC RIN RN RSN TAPA/TAPhA TM TTH TU ZPapEpig/ZPE ZRVI

xiii

The Prosopography of the Later Roman Empire, vols. 1 and 2. Cambridge, 1971–1980. Patrologia orientalia Revue belge de numismatique Georg Wissowa et al. (eds). Paulys Real-Encyclopädie der classischen Altertumswissenschaft. Stuttgart, 1894–1963. Revue des études anciennes Atti dell’accademia nazionale dei Lincei, Rendiconti, Classe GLVFLHQ]HPRUDOLVWRULFKHH¿ORORJLFKH 5LYLVWDGL¿ORORJLDHGLLVWUX]LRQHFODVVLFD Rheinische Museum für Philologie H. Mattingly et al. Roman Imperial Coinage. London, 1923–. 5LYLVWDLWDOLDQDGLQXPLVPDWLFDHVFLHQ]HDI¿QL Revue numismatique Revue suisse de numismatique Transactions of the American Philological Association Travaux et mémoires Translated Texts for Historians. Liverpool, 1985–. Texte und Untersuchungen zur Geschichte der altchristlichen Literatur. Leipzig and Berlin, 1883–. Zeitschrift für Papyrologie und Epigraphik Zborknik radova Vizantinoškog instituta

I THE DATES AND EDITIONS OF EUSEBIUS' CHRONIC! CANONES AND HISTORIA ECCLESIASTICA THE earliest evidence we have for the existence of Eusebius' now lost XpovtKoi Kav6vEs (Chronici canones or Chronological Tables) comes from other works of Eusebius: the Historia ecclesiastica (HE) (1.1.6), the first edition of which is variously dated between pre-293 and 3 I 3 (see below); the preface to book six of the General Elementary Introduction, of which four books (6-9) survive under the title Eclogae propheticae (PG 22. 1023A), dated 303/3 12; 1 and the Praeparatio Evangelica ( 10.9. II), dated c.314/318. This early evidence demonstrates that there must have been an edition earlier than the one of 325, where the universal testimony of the surviving witnesses places its conclusion. 2 Until now, there has been no solid evidence to suggest when any such putative first edition may have been produced and consequently there has been much debate and discussion.

The following are cited by short title only: Barnes, 'Editions'= T. D. Barnes, 'The Editions of Eusebius' Ecclesiastical History', GRBS 21 (198o), 191-20I. Barnes, C and E=T. D. Barnes, Constantine and Eusebius (Cambridge, Mass., 1981). Louth, 'Date'=Andrcw Louth, 'The Date of Eusebius' Historia Ecclesiastica', JTS, NS, 41 (1990), III-2J. 1 The tenth and last book of the General Elementary Introduction-surviving as the Commentary on Luke-must date after 309; see D. S. Wallace-Hadrill, 'Eusebius of Caesarea's Commentary on Luke: Its Origin and Early History', HTR 67 (1974), 6J. 2 See Alden A. Mosshammer, The Chronicle of Eusebius and Greek Chronographic Tradition (Lewisburg, PA, 1979), 38, 61, 62-63, 75· For the conclusion in year twenty of Constantine (=AD 325), see e.g. Eusebius, Chronographia (Greek: John Anthony Cramer, Anecdota Graeca e Codd. Manuscriptis Bibliothecae Regiae Parisiensis, vol. 2 (Oxford, 1839; repr. Hildesheim, 1967), 160.8-9; Armenian translation: Josef Karst (ed.), Eusebius Werke 5: Die Chronik aus dem Armenischen Obersetzt mit textkritischem Commentar (GCS 20; Leipzig, 191 1), 62.3-5); Chronici canones (Latin translation of Jerome: Rudolf Helm (ed.), Eusebius Werke 7: Die Chronik des Hieronymus 3 (GCS 47; Berlin, 1984)), 6.17-8, 231f; Chronicon miscellaneum ad annum Domini 724 pertinens (Syriae epitome of the Canones; CSCO 4, Chron. min. 2: Scriptores Syri, series 3, tomus 4, versio, by J.-B. Chabot), 100.22, 32-3; Samuel Aniensis, Summarium temporum, PG 19.665; Chronicon Paschale (Ludwig Dindorf (ed.), CSHB 16 (Bonn, 1832)) s.a. 325, pp. 526.5-6; 527.2-5; and James of Edessa, Chronicon (Syriac continuation of Eusebius; CSCO 6, Chron. min. 3: SS series 3, tomus 4, versio, by E. W. Brooks), 199, 200, 203, 204, 205, 209, 214.

I 472 I The communis opinio is that the first edition was completed in or around 303. 3 This view was first popularized by an influential article in Pauly-Wissowa by Eduard Schwartz, who believed that the Canones had to date before the Eclogae propheticae, but did not believe that Eusebius could have written such a work during the persecution. He therefore stated that Eusebius had written it or at least collected his materials before 303. 4 The recent currency of 303, however, chiefly depends on an article written by D. S. Wallace-Hadrill in 1955, which was based on the earlier hypotheses of Joseph Karst, editor of the Armenian translation of the Canones. 5 Wallace-Hadrill accepted Karst's argument that the Armenian translation represented the first edition of this work and that the terminal date of the Armenian Canones, Year 16 of Diocletian ( = 300), was thus the concluding date of the first edition. He also attempted to buttress Karst's hypothesis with other evidence for a visible 'joint' between the first and second editions. He cites from Jerome's translation three additions and alterations to the Canones that '[cluster] round the year 303' (pp. 249-50). 6 However, though these probably are all later additions and alterations, there is no reason why such 'rewriting in

*

3 For this date, see, for example, Otto Bardenhewer, Geschichte der altkirchlichen Literatur 3 (Freiburg, I923), 248-49; Kirsopp Lake, Eusebius. The Ecclesiastical History I (Loeb Classical Library; New York, I926), xvii; Johannes Quasten, Patrology 3 (Utrecht/Antwerp, I96o), 3 I2; Berthold Altaner, Patrology (New York, I96I), 264; Mosshammer (cit. n. 2), p. 32; R. M. Grant, Eusebius as Church Historian (Oxford, I98o), I; Johannes Karayannopulos and Gunter WeiB, Quellenkunde zur Geschichte von Byzanz (324-I453) 2 (Wiesbaden, I982), 244; Frances M. Young, From Nicaea to Chalcedon (London, I983), 5; W. H. C. Frend, The Rise of Christianity (London, I984), 457, 477, 478; Robin Lane Fox, Pagans and Christians (London, I986), 6o6; C. Curti in Angelo Di Berardino (ed.), Encyclopedia of the Early Church I (New York, I992), 299; and others cited by Barnes in 'Editions', p. I93, and C and E, p. 34I n. 67. 4 RE 6. I (I907), p. I376. 5 D. S. Wallace-Hadrill, 'The Eusebian Chronicle: The Extent and Date of Composition of its Early Editions', JTS 6 (I955), 248-53 (repeated in his Eusebius of Caesarea (London, I96o), 43) and Karst (cit. n. 2), pp. xxx-xxxiii. 6 These are the notice concerning Constantine's accession in the fourth year of the persecution ( = 306) under Year I9 ( = 303), the alteration of the month of the inception of the persecution from April to March under Year I9, and the reference to the martyrdom of Peter of Alexandria in the ninth year of the persecution (t25 November 3II) under Year I9 (though Wallace-Hadrill did not know that this is Jerome's error: in Eusebius' original it was dated to Year I7. For a reconstruction of Eusebius' original text for these years, see my paper 'The Chronici canones of Eusebius of Caesarea: Chronology and Content, AD 282-325', which is nearing completion. These entries may appear in Year I9 of Diocletian in Jerome, but only one actually has anything to do with 303.

I 473 the light of later knowledge' could or should have occurred only at the end of the first edition. The key point is, in fact, that they all relate to later knowledge concerning the persecution, which began in 303. These entries, therefore, have no bearing on the date of the first edition. Wallace-Hadrill also notes that Eusebius' list of the bishops of apostolic sees stops in 302 (p. 250). This is true, but apart from Rome (for which information would have been difficult for Eusebius to obtain during the persecution), no further bishops were ordained in Antioch, Jerusalem, or Alexandria until about 312/3 (Vitalis, Macarius, and Achillas, respectively). Following Wallace-Hadrill's argument any date as late as 312/3 is therefore possible. But the episcopal lists in both the Canones and the HE purposely cease with the beginning of the persecution, not the end of the first edition of the Canones. Eusebius explicitly says this in the HE (7.32.32 and 8 pref.), though he does not explain why. For some reason apostolic succession was no longer important during the persecution or in its aftermath. If the end of the list simply marked the end of the first edition, there is no reason why Eusebius should not have continued the list in the later editions of both works, especially the HE. Once again, the crux is the beginning of the persecution, not the end of the first edition. Finally, Wallace-Hadrill points to twelve differences between the Armenian translation and Jerome's translation (pp. 251-52), claiming that these arise because each translation represents a distinct edition-the Armenian the first edition, Jerome's Latin the third edition. The argument is irrelevant, however, since none of these twelve items relates to 303. It is further flawed by the fact that the Armenian translation is not a different edition from Jerome's and is not complete as Karst believed; it is simply a defective translation of a composite Armenian/Syriac version of the same 325 edition as Jerome's.' The differences that WallaceCHRONIC! CANONES/HISTORIA ECCLESIASTICA

7 In the Armenian translation of the Chronographia, which is Eusebius' lengthy introduction to his sources and establishment of the individual chronologies for the Canones, there is a note that mentions the twentieth year of Constantine (Karst (cit. n. 2), p. 62.3-5, which is the same as the Greek fragment printed by Cramer (cit. n. 2), p. 160.8-9) and an identical Armenian translation was used by Samuel Aniensis in the late-twelfth century for his Armenian chronicle and he records the conclusion of Eusebius' chronicle in 325 as well (PG 19.665). Furthermore, the last entry in the Armenian translation (on Hermon of Jerusalem) appeared under Year 17 of Diocletian in Eusebius' Greek original (see my paper cited in n. 6, above) and has mistakenly slipped back a year, thus avoiding the oblivion shared by the rest of the text after Year 16. Karst claimed that the Armenian translation had been contaminated by the edition of 325. See Mosshammer (cit. n. 2), pp. 59-60, 75·

I 474 Hadrill points out (and many others that he does not mention) arise simply because the Armenian version is not a complete or accurate translation; its various translators and redactors omitted and altered textual material and chronological markers through wilful error, carelessness, or lack of interest (for two examples, see nn. 19 and 27, below). Unfortunately, on occasion Jerome made mistakes as well. 8 The major problem with Wallace-Hadrill's argument, however, is that he claims to accept Karst's conclusions as the foundation of his own argument, yet redefines them: Karst's hypothesis was that the first edition ended in Year 16 of Diocletian, that is 300, but Wallace-Hadrill changes this to 303 (which for some reason he insists on labelling 'Diocl. 18', when it is in fact 'Diocl. 19'). Karst's entire case rested on the supposition that the Armenian translation was complete. Wallace-Hadrill abandons that supposition, stating that the first edition 'did not extend far beyond the mutilated end of the Armenian text' (p. 250), but in so doing he unwittingly abandons Karst's entire hypothesis and hence the foundation of his own: if the Armenian translation does not end in 300, the evidence shows that it must have concluded in 325 and it is consequently irrelevant to Wallace-Hadrill's argument for 303. He tries to paper over the gap between 300 and 303 with a rather nonsensical note-' If Diocletian became emperor in 284, his 16th year is 300, though the Arm. Chron. aligns the regnal years with the Olympiads so as to make it 303' (p. 248 n. 8) 9 but the fundamental contradiction remains. There is, therefore, no valid evidence that Eusebius concluded the Canones in or just before 303. More recently T. D. Barnes has come out strongly in favour of 277 (Year 2 of Probus) as a terminal date for the Canones and the early 290s as the date of composition, though this view has not gained widespread acceptance. This date depends chiefly upon his early dating of Eusebius' Onomasticon and HE, both of which must have been written after the Canones. 10 The specific terminal date of 277 for the Canones follows an earlier suggestion by Rudolf Helm. 11 The sole direct evidence for this conclusion is the fact 8 For the problems with the Armenian translation, see Mosshammer (cit. n. 2), pp. 6o-63, 73-79. For Jerome, see R. W. Burgess, 'Jerome and the Kaisergeschichte', Historia 44 (1995), 355 n. 31, and my forthcoming paper cited inn. 6, above. The overall accuracy of Jerome is confirmed by comparison with the Syriac traditions and other Greek witnesses. 9 In his book (cit. n. 5) he strays even further from his own argument and the truth: 'the sixteenth year of Diocletian ... in Eusebius' dating is 303' (p. 43). 1 For this, see below, nn. 17 and 29. 11 'Editions', p. 193, and C and E, pp. uo-11, 113, and 146.

°

I 475 that it is in this year that one finds a synchronism of five local eastern calendars-those of Antioch, Tyre, Laodicea, Edessa, and Ascalon-with Year 2 of Probus. 12 The first problem with the date of 277 is that Eusebius would only have been at most seventeen years old when he wrote the chronicle. 13 This is virtually impossible and Barnes actually posits composition almost fifteen years later, in the early 29os/ 4 yet such a large gap between the date of composition and the conclusion of the work is most implausible. 15 Barnes explains the gap by claiming that Eusebius ended his chronicle in 277 as a compliment to Anatolius, Bishop of Laodicea (Canones 223i), whose famous Easter canon either began or ended in that year, but he does not explain the connection between Anatolius and the five local calendars noted by Eusebius. Unfortunately, there seems little reason why Eusebius would have ended the Canones, a work of universal Christian history and chronology, with such an obscure and irrelevant set of local synchronisms, simply because a famous Easter canon began or ended CHRONIC! CANONES/HISTORIA ECCLESIASTICA

12 On this synchronism, see the comments of Mosshammer (cit. n. 2), p. 75; Grant (cit. n. 3), pp. 7-9; and Eduard Schwartz, Eusebius Werke 2. Die Kirchengeschichte (GCS 9.3; Leipzig, 1909), ccxlvi--ccxlvii. Mosshammer claims that this summary 'has no parallel except at the very end of the work', but this is a misrepresentation of the final supputatio, which records the number of years elapsed from seven key dates in history to the conclusion of the chronicle. It is not in any way similar to this list of dates. 13 Born around 260/265; see Barnes, C and E, p. 277. 14 Barnes actually believes that c.293 is the terminus ante quem (evident from C and E, pp. 1 ro-1 1, and The New Empire of Constantine and Diocletian (Cambridge, Mass., 1982), 215, discussing boundary changes to Palestine c.293 that he uses to date the Onomasticon (see n. 17, below)), but he is unduly vague about the exact date of composition: 'he had completed the Chronicle by ... ca 295', 'Editions', p. 193; 'before the end of the third century', C and E, p. 111; 'at least a decade earlier [than 303]', p. 113; 'before 300', p. 277; 'c.295', p. 346 n. ro; and 'the 290s', 'Scholarship or Propaganda? Porphyry Against the Christians and its Historical Setting', BICS 39 (1994), 59· 15 Eusebius' youth in c.28o is noted with some surprise by Brian Croke ('The Origins of the Christian World Chronicle', in B. Croke and A. M. Emmett (eds.), History and Historians in Late Antiquity (Sydney, 1983), p. 128 n. 4), who seems unaware of Barnes' argument for later compilation and assumes that the Canones would have been completed within a few years of the date of its conclusion (see also Brian Croke, 'Porphyry's Anti-Christian Chronology', JTS, NS, 34 (1983), 171 and 184). In a later article ('The Era of Porphyry's Anti-Christian Polemic', JRH 13 (1984-85), 10 n. 53) he suggests the 28os since 'the Chronicle is not the work of a novice'. Croke's reaction is to be expected since it is unheard of for a chronicler to conclude an original chronicle ten to fifteen years before the time of writing. Andrew Louth ('Date', p. 121), with reference to the HE, which Barnes dates to the same period as the Canones (see III, below), states, 'Eusebius was then in his twenties, or even passing from his mid-teens to his mid-twenties: one wonders if he could really have read as much as the Historia Ecclesiastica presupposes by then.'

I 476 in that year. It makes no sense for Eusebius to have charted the history of the world from the birth of Abraham in 2016 BC, only to ignore the history and chronology of the most recent fifteen years of Christian growth and advancement as a compliment to the author of merely one of what must have been many competing Easter canons. There is no connection between the two works or the authors, apart from the fact that Eusebius admired Anatolius (cf. HE 7·32.13-21, where Eusebius quotes from his works, including the canon, simply as an example of Anatolius' wide learning), and even if there were, the synchronization would still be a tribute to Anatolius whether the Canones ended there or not. 16 Year 2 of Probus cannot therefore have been the conclusion of the first edition. Barnes also offers a series of lesser interlocking arguments in support of a date of composition in the early 290s but these do not stand up to careful scrutinyY

II Unfortunately, an edition concluding in 303, 300, or 277 suffers from another more serious problem and that concerns the indications of the date of composition derived from Eusebius' own chronology. The key lies in two examples of obvious, and in one case bizarre, tampering with the regnal year chronology of Carus, Carinus, and Numerian, and Diocletian. Eusebius assigns the reign of Carus and his sons only two years instead of three, a peculiar mistake for a contemporary reign that Eusebius should have known well. 18 Even more peculiar, he omits the single regnal year of Constantius I ( = 305), but attributes to the preceding Year 20 of Diocletian ( = 304) two Years of Persecution (2 and 3, March/April 304 to March/April 306), two Years of Abraham 16 There is also the observation made by Barnes (C and E, p. I I I) and R. M. Grant (cit. n. 3, pp. 7-9), who point to the appearance of the eighty-sixth Jubilee in the same year of Probus, but what relevance this could have to the calendar synchronization is unknown. 17 These chiefly concern the argument that the Canones and the HE were written before the Onomasticon, dated by Barnes to c.293, which is demonstrably false, since the dedication to Paulinus alone dates the work to c.3I3-24. Barnes' greatest impediment is that he accepts as ultimately Eusebian passages in Jerome's Latin translation of the preface to the Onomasticon that do not appear in the original Greek. A number of other problems are discussed by Louth, 'Date', pp. I I 8-20. 18 Since Carus became emperor in the autumn of 282 and Diocletian sole emperor around the summer of 285 (a period of just under three years), Carus and his sons should have been allotted three regnal years (for 282, 283, and 284). For Eusebius' regnal years, see below. That Eusebius knew that they did indeed reign 'for not three whole years' is demonstrated by HE 7.30.22. The regnal years in the HE appear to derive (often in an extremely careless manner) from either the Canones or the same sources as the Canones.

I 477 (232I and 2322), and two Olympiads (271.I and 2). Though Constantius' death appears in the second half of this split regnal year (Persecution 3, 2322 Abr., and Olymp. 27I .2), his regnal year total, twelve years, only counts Year 20 of Diocletian once (his accession is dated to Year 9 of Diocletian, though this was altered by Jerome as was the regnal year total in consequence). This doubling up of Year 20 appears in Jerome (228d-g) and is confirmed by the Years of Abraham in Pseudo-Dionysius, who assigns 232I Abr. (p. I I2.27) to Jerome's entry 228d, the retirement of Diocletian, and 2322 Abr. (p. II3.I) to entry 228g, the death of Constantius, both in Year 20 of Diocletian. 19 It is also confirmed by the total number of years of Abraham noted in Eusebius' supputatio as preserved in the Chron. 724 (p. I00.23): though there are only 297 regnal years covered between Year IS of Tiberius and Year 20 of Constantine (325-28 = 297), Eusebius lists 298 Years of Abraham (2342-2044=298). This attribution of two calendar years (Years of Persecution), two Years of Abraham, and two years of an Olympiad to a single regnal year, which itself represents a single calendar year, is unique in the chronicle. An explanation of this tampering requires an understanding of Eusebius' use of regnal years and his regnal year chronology. 20 Throughout the early imperial period, starting with Julius Caesar's sole rule in 48 BC, Eusebius' regnal-year chronology is almost perfectly accurate. Each regnal year is treated as the equivaCHRONICI CANONES/HISTORIA ECCLESIASTICA

19 Jerome's Years of Abraham elsewhere generally agree with those in PseudoDionysius (J.-B. Chabot (trans.), Chronicon Pseudo-Dionysianum uulgo dictum, CSCO I2I: SS series 3, tomus I, versio; Louvain, I949), except in a few places where scribal corruption or simple copying errors are involved. Socrates (HE 1.2) mentions the death of Constantius (in Eusebius =Year 20 2 ), but does not realize that there are two Olympiads in the one regnal Year (they are only marked every four years), and so quotes the Olympiad for Year 20\ 271.1. This Olympiad agrees with that for Year 20 1 in Jerome. Chron. Pasch. (5I8.II, with 5I4.I8 and 5I9.3; and 524.I7, with 524.9) assigns Year I of Constantine to Olymp. 271.3 and Year 20 to 276.2, just as in Jerome (though the Olympiads are off by one for the reign of Diocletian because the Chron. Pasch. assigns a correct three years to Carus and his sons). These agreements show that the Armenian translation is not an accurate account of the chronological relationship among the regnal years, Years of Abraham, and Olympiads in Eusebius: e.g. Year 201 in Jerome is 232I Abr. and 01. 271.I; in Ps-Dion. is 232I Abr.; in Socrates. is 01. 271.I; in Samuel Aniensis (which is based upon the Armenian translation) is 01. 271.4 (col. 663); and in the Armenian translation would be 2323 Abr. and 01. 271.4, if it went that far. 20 It should be noted that Jerome completely altered Eusebius' chronology for the reign of Constantine to actually count the doubled regnal Year 20 of Diocletian as two calendar years ( = 305 and 306), and Helm's marginal accounting of years follows Jerome's sequence not Eusebius'. Eusebius counted Year I of Constantine as 306 and the following years in sequence up to Year 20 in 325. For the details of this, and what follows below, see my paper cited in n. 6, above.

I 478 lent of a full calendar year, whatever calendar it was that he was using. 21 In reality, of course, an emperor's first and last regnal years were only part of a calendar year. When regnal years are treated only as indivisible full years some accommodation must be made. Eusebius did this by placing the death of Emperor A and the accession of Emperor B a year early, so that Emperor B's first regnal year would correspond to the true calendar year of his accession. Each emperor's accession immediately follows the death of his predecessor and the first regnal year of an emperor immediately follows his accession. Thus the death of Emperor A and the accession of Emperor B occur in the same year, the last year of Emperor A, and Emperor B's first regnal year then usually begins immediately. Eusebius was able to maintain a perfect chronology throughout the early part of his imperial history because of the accurate records of the lengths of imperial reigns and detailed information concerning the years in which the emperors came to the throne and died. \Vithin the first z6o years he errs only once, but that error was deliberate and makes no difference to the overall sum of regnal years. 22 However, once he advanced his chronology into the third century, he made three serious errors that he did not fully compensate for and that therefore disrupted his entire chronological sequence: he assigned Caracalla seven regnal years instead of six, Philip seven instead of five, and Decius one instead of two.Z 3 By the time he reached the accession of Carus in 282 his chronology was consequently two years ahead of itself: Year I of Carus was the equivalent of 284 instead of 282. The chronological tampering with the reigns of Carus and his 21 Eusebius derived his accession dates for the emperors from Caesar to Caracalla, at least, from an Olympiad chronicle that equated each Olympiad with a Seleucid/Macedonian year that appears to have begun in the middle of September or perhaps on I October. 22 All the emperors between Caesar and Caracalla inclusive, and then Constantine, have their accessions placed in the correct Olympiad/Seleucid year, with the exception of Augustus, whose accession was delayed to 45 BC so that the famous murder of Caesar could appear in the correct 44 BC. If the calendar on which Eusebius' Olympiad dates were based did begin on I October (as it did in Antioch, for instance, the standard Eastern calendar), the accession of Nerva (I8 Sept 96) would have been placed one year too late (97) as well. 23 The same figures for the first two appear in HE 6.2I.I (actually seven years and six months!) and 39· 1. In 7. 1. I he says that Decius reigned 'for not two whole years', which is correct (about a year and eight months). The one year and three months of the Canones must therefore be a misreading of the source that he read more carefully for the HE. The accessions of the following emperors are consequently late: by one year, Macrinus, Elagabalus, Severus Alexander, lVIaximinus, Gordian III, Philip; by three years, Decius; by two years, Gallus and Volusianus; Valerian and Gallienus; Claudius; Aurelian; Tacitus; Probus; Carus, Carinus, and Numerian; by one year, Diocletian.

I CHRONIC! CANONES/HISTORIA ECCLESIASTICA 479

sons, and Diocletian is serious and complicated. Why would Eusebius have fiddled these regnal years in such a bizarre and obvious manner? The answer arises from the consequence of the tampering: by cutting Year 3 of Carus, Carinus, and Numerian, and Year I of Constantius, Eusebius was able to shed in a short space of time the two extra regnal years that he had accumulated throughout the third century (Year 7 of Caracalla plus Years 6 and 7 of Philip minus Year 2 of Decius). In Year 2 of Carus and his sons Eusebius' overall regnal year chronology was still out by two years, having been one, two, or three years ahead of itself for almost seventy years, yet within twenty-one years, by Year I of Constantine, it was once again synchronized with the overall chronology of calendar years, as it had been in the first and second centuries. This chronological tampering can only be explained by the hypothesis that Eusebius was determined to conclude his chronology with a correct overall correlation between calendar and regnal years (which is, of course, the whole point of compiling such a chronology). If we look at his entire imperial chronology from Year r of Caesar (=48 BC) to Year 20 of Constantine (=AD 325) we can see that Eusebius assigns 373 regnal years to 373 calendar years. This synchronization is valid back to Year I of Constantine ( = 306, thus 354 regnal years over 354 calendar years). Any further back and the sequence is disrupted, by one year for Diocletian and by one or two years for much of the rest of the third century. Only in Year 6 of Caracalla does it come back into synchronization. For example, as I noted above, in Year 2 of Probus Eusebius notes the synchronization of five local calendars. In the case of two of these, Antioch and Edessa, he has in fact noted the date of the beginning of each era in its correct place. The beginning of the era of Antioch is noted at I56b in I969 Abr. ( =48 BC). Eusebius states that Probus 2 (2295 Abr.) is year 325 of the era of Antioch (as it is, 277 + 48), but there are 327 years of Abraham between the two notices since Probus 2 is the equivalent of 279, not 277 as it should be. He also notes the beginning of the era of Edessa at I26h in 1706 Abr. ( = 3 I I BC, i.e. the well-known Seleucid era), stating that Probus 2 is year 588 of the era of Edessa (as it is, 277 + 3 I I), but there are 590 years of Abraham between the two notices. 24 In the HE (7.32.32) he states that there were 305 years between the birth of Christ and the inception of the Great Persecution (=March/April 303, assigned to 2320 Abr., which is 304), but since he places the incarnation in Year 42 of Augustus, 24 These show that the correlation of these five local eras with Year z of Probus was copied by Eusebius from another work.

I 480 which is 20I 5 Abr. ( = 2 BC, 2320-2oi 5 = 305), this is one year too many (302 years+2 years=304 years, or 23I9-20I5=304). 25 As I noted above (n. 2I), for the early part of his imperial chronology, from Caesar to Caracalla, Eusebius had a source that dated events by Olympiads and this provided him with the correct year of each emperor's accession. Thus from 48 BC to AD 2I6 his imperial chronology is perfectly accurate. From the reign of Caracalla, however, he had to rely solely on the length of each emperor's reign in years and months to construct his regnal year chronology. To this he must have added a local era, such as the years of Antioch or Edessa, that would have provided known contemporary dates leading back to a beginning fixed at some accurately established point in the past. It was while he was using these sources that his chronology got ahead of the correct calendar years since he did not know in what year anyone became emperor, he only knew how long each was emperor (and this information was often inaccurate). He obviously knew his contemporary chronology and was easily able to equate current regnal years with the years of his local calendar and then work them backwards with perfect accuracy. We can see from his preface and supputatio that he did indeed work his chronologies both backwards and forwards (Jerome, IO-I8, 250). It was only when his inaccurate noncontemporary history, worked forwards, met his accurate contemporary history, worked backwards, that he ran into difficulty: he faced an overlap of two years. It is Year 20 of Diocletian that marks this final 'seam' between Eusebius' accurate contemporary chronology and his inaccurate non-contemporary chronology. He seems not to have been able to calculate where the errors of his earlier chronology were and he could not simply cut two years from the end of Diocletian's reign. He made his first cut in the reigns of Carus and his sons. This left one year unaccounted for at the end of Diocletian's reign, and so rather than cut Diocletian's total, he ingeniously opted to cut the single regnal year of Constantius (I May 305 to 25 July 306). In this way Constantine's regnal years could start with Year I accurately associated with the equivalent of 306 right after Year 20 of Diocletian, even though Year 20 of Diocletian was the equivalent of 304 (since Year I= 285). The solution was almost perfect. Unfortunately for Eusebius, at exactly this point he was dealing with a subsidiary 'local' chronology, the Years of Persecution, which extended from Year I9 of Diocletian. When 25 I count two years because Eusebius would have dated the Nativity to 6 January 2 nc not 25 December, which is a western tradition.

I CHRONIC! CANONES/HISTORIA ECCLESIASTICA

481

he cut Year I of Constantius, he also cut Year 3 of Persecution. In order to maintain his overall accounting of regnal years, but still be able to take account of the cut regnal/persecution year locally, he added the events of Constantius' sole year as emperor (the death of Constantius and the accession of Constantine) and the marker for Year 3 of Persecution to Year 20 of Diocletian, which was already Year 2. The doubled years of Abraham and Olympiads are a later compilation error created in the final draft by mistakenly counting the Year 3 of Persecution as a regnal year, hence the extra Year of Abraham noted above (p. 477). The Years of Persecution consequently run correctly in spite of the excised regnal year, there being ten Years of Persecution between Year I9 of Diocletian (303) and Year 7 of Constantine (3I2), even though there are only nine regnal years involved (2 + 7). Confirmation that this doubling up of dates is in fact an attempt by Eusebius to maintain what he perceives to be the correct chronology comes from a similar chronological 'fudge' to be found at Jerome, 105-106. From his research Eusebius knew that the rebuilding of the Temple took place in the second year of Darius and Olympiad 65.I (Chronographia, pp. 57-59 (Karst); Canones, IO. I2-3, I8.3-5, and 105ac). Unfortunately his Persian chronology is one year short, so the second year of Darius actually ends up in Olymp. 64.4. To rectify the situation he repeats Darius' second year again in Olymp. 65.I, even though the regnal years for the parallel kingdoms advance one ye,ar (Tarquinius Superbus of the Romans from 27 to 28 and Amyntas of the Macedonians from 33 to 34· Next to the first second year he adds the following note, Ideo secundus annus bis scribitur, quia unus annus in magorum fratrum VII mensibus computatur (Helm, I05a"'; cf. I04a.22-26). This is essentially the reverse of what I have described above under Diocletian, an expansion of regnal years rather than a contraction. If the Canones had been compiled at any date before 306 the chronology at the end of the work would have come back into synchronization around the date of composition, as Eusebius tried to match contemporary chronology worked backwards with noncontemporary chronology worked forwards. Year 2 of Probus should be the equivalent of 277 (following Eusebius' method of placing Year I in the year of accession), yet it is in fact the equivalent of 279. The accession of Carus is also two years late, 284 instead of 282. The reigns of Probus and Carus are therefore treated just as vaguely and inaccurately as any of the earlier thirdcentury reigns; there is no evidence of contemporary compilation around Year 2 of Probus. It cannot be the concluding point of the first edition. The reign of Diocletian is also out by one year, so

I 482 neither 300 nor 303 can be the concluding point of the first edition. We have seen above both the lengths that Eusebius went on to achieve correct contemporary chronological synchronization at the end of his work and his ability to calculate accurately and record almost 370 years of imperial chronology. We have no grounds for assuming that such zeal for chronological accuracy and the accompanying skill were lacking in the first edition. The first edition of the Canones must therefore date after 306, since that is obviously the seam between contemporary and non-contemporary history. In the tangled web of controversy concerning the dating of the Canones and the HE, this chronological observation is the strongest evidence yet advanced. And if the Canones must date after 306, then the HE must as well.

III Eusebius states in HE 1. 1.6 that his Chronici canones was completed before the HE, though he gives no indication of the lapse of time. The nature of the comment suggests that it had been long enough for the Canones to get into general circulation, since he is countering a potential criticism of his new work, that it covers the same ground covered by the Canones. He states that the Canones was merely an eTTtTop.fJ, while the HE had a narrative that was 7TA1JpEaTaT7J. It is obvious that he has used the Canones as a source, however summarily, but he has revised some of the chronology and content as a result of his more extensive and careful reading, one supposes, though he failed to emend the text of the Canones in accordance with it, even in his later editions, probably because of the daunting nature of such a task. This is an important observation, because it disproves any theory that holds that there were major differences between editions of the Canones. 26 A further example confirms this conclusion. In the Canones he states that there were 406 years between the first Olympiad (1241 Abr.) and the capture of Troy (835 Abr.) (1 1.8,

26 See Mosshammer (cit. n. 2), pp. 75: 'a major change in format between the two versions is not likely. Eusebius had only to add a few pages to the Chronicle, not rework the whole', and 6r, quoting J. K. Fotheringham, 'Aliud est producere, aliud redigere'. See Croke ('Era', cit. n. 15), pp. 12-13, esp. p. 13: 'Given the tedium and complexity of copying a chronicle like that of Eusebius, it is scarcely likely that the whole chronological frame-work set out in the Chronographia was reworked when the point of the second edition was simply to bring the work up-to-date, that is by expanding the canons'.

I CHRONIC! CANONES/HISTORIA ECCLESIASTICA

483

=406)/7

yet I3; 6oac; 250.IO, I4 [IS6I-II55 =406], and I24I-835 in the Praeparatio Evangelica (written between c.3I4 and 3I8) he twice gives the figure as 408 years (I0.9.6 and 7). If, as some claim, Eusebius had revised the text of the Canones for the edition of 325, he obviously would have changed his chronology of 406 years to 408 years as part of that revision, to accord with his new calculations. He obviously did not, which indicates that the chronology of the 325 edition remained unrevised in the light of new calculations made in the 3IOs. As we shall see below, there is no evidence that he undertook extensive revisions of the HE either. There are a number of hypotheses concerning the date of the first edition of the HE, the common being that it originally concluded with book seven (c.303, but before the persecution, thus at the same time as the Canones), with book eight (c.3I I), or with book nine (3I3/I4). 28 As I have noted above, T. D. Barnes argues for a first edition before c.293, concluding, like the chronicle, around 277, near the end of book seven. In his view the HE represents a universal history of the church only down to the late 270s, which must therefore mark the end of the first edition. 29 It is these early dates especially that must concern us now, for solid evidence dating the HE before 306 would seriously undermine the date for the Canones that I am proposing. The manuscripts reveal traces of editions completed in 3I3/4, 3 I 5/6, and 324/5, and a modification of the 324/5 edition in 326 to remove Crisp us (this latter only in the Syriac translation). 30 The evidence for these editions is chiefly the textual variants arising from later emendation in the early Greek manuscript tradi27 The 405 years stated at 86a" is a scribal error on Jerome's part. The Armenian translation gives 405 years between the two events because it antedates the start of the Olympiads by one year (I240 Abr. instead of I24I). Jerome's chronology is confirmed by other witnesses. 28 For various supporters of these views, see Barnes, 'Editions', pp. I9I n. 2 and I99, and C and E, pp. 346-47 n. Io; Quasten (cit. n. 3), pp. 3I5-I6; WallaceHadrill (cit. n. 5), pp. 39-43; Lake (cit. n. 3), pp. xix-xxiii; and Louth, 'Date', pp. II2-I3, II4-I5, and I22-23. To these can be added Grant (cit. n. 3), pp. I4-I5, for a date not much earlier than 303. Grant's reasons for dating the work to this period are too subjective to be of assistance in this analysis ('an earlier date rather than a later one would allow adequate time for the changes within the first seven books which we hope to establish', p. I5). With regard to these changes, Barnes rightly concludes, 'I do not believe that Grant has established [his conclusions] satisfactorily', C and E, p. 346. See also T. D. Barnes, 'Some Inconsistencies in Eusebius', JTS, NS, 35 (I984), 470-75. 29 'Editions', pp. I99-20I; C and E, pp. III, I28-29, I45-47; and personal communications. 30 This was first established by Schwartz (cit. n. I2), pp. xlvii-cxlvii, and is described by Quasten (cit. n. 3), p. 3I5; Barnes, 'Editions', pp. I96-98; Grant (cit. n. 3), pp. IO-I3; and Louth, 'Date', pp. III-I2, II3-I4.

I 484 tion: minor tampering for political reasons in most places and the rewriting of book eight in 3 I s/6 to compensate for the removal of the short recension of the Martyrs of Palestine (see Appendix I). There is no trace in the manuscripts of any edition earlier than 3 I 3, especially in books six and seven, where the supposed differences between the earlier and later editions were great (see below), even though Barnes' putative first edition of before c.293 was in circulation for almost twenty years and that of c.303 for ten years, certainly long enough to leave some trace in a tradition that can otherwise distinguish editions completed fewer than five and ten years apart. This lack of manuscript evidence makes an edition before 3 I 3 most doubtful. 31 Adding to this doubt are the obvious later references scattered throughout books one to seven. 32 None of these references suggests a date later than the end of book nine, which corresponds to the manuscript evidence, an important agreement that seems to have been overlooked. A number of these references indicate other works that Eusebius had written, such as the Eclogae propheticae (303/3I2; 1.2.27) and the Life of Pamphilus (3I0/3II; 6.32.3; 7.32.25), or works of others that did not appear until later, such as the forged Acts of Pilate (probably c.3 I I; 1.9.3-4, I 1.9), the Doctrine of Addai (c.3oo; 1. I 3), and Porphyry's Against the Christians (c.300 (or perhaps c.275); 6. I9. 2-I I). Nowhere does Eusebius refer to works written after 3I3. Louth notes that Eusebius' account of Origen, which takes up much of book six (roughly I-6, 8, I4-I9, 2I, 23-28, 30, 32-33, 36-39 of 46 chapters), refers three times to Pamphilus and Eusebius' Defence of Origen (6.23.4, 33.4, 36.4) and almost certainly derives from it, though it was not written until 308/3 I o. 33 If this material had not originally existed, book six would be much shorter indeed and would lack its central unifying focus. In books one to seven Eusebius also makes reference to later events, especially the Great On this, see also Lane Fox (cit. n. 3), p. 607. Barnes provides lists in 'Editions', p. 201 n. 28, and C and E, pp. 146, 346 n. ro, and 355 nn. r66-67, 170, 172: r.r.2, 2.14-16, 2.27, 9.3-4, rr.9, 13; 4.7.14; 6.19.2-15, 23.3-4, 32.3, 33-4, 36-4-7; 7.18.3, 30 index and chapter heading, 30.22, 31, 32.1-4,22-32. To these can be added 7.1r.26 (on the persecution) and 7.30.21 (which refers ahead to 8.r.7-9). Yet Barnes claims that Eusebius only 'slightly retouched' the first seven books when he created the second edition (C and E, p. 149). As a general, though not invariable, principle, I agree with G. A. Williamson in the preface to his Penguin translation of the HE: 'in the absence of textual evidence that they are afterthoughts we ought to treat all references to late events as proof of late writing' (p. 21). 33 'Date', pp. 121-22. See also Lane Fox (cit. n. 3), pp. 6o7 and 774 n. 33· See Wallace-Hadrill (cit. n. 5), pp. 160-65, who derives all or parts of I-3, 8, 15-16, r8-19, 23, 28, 30, 33, 36, 37, and 39 from the Defence of Origen. 31

32

I 485 Persecution: 1.1.2; 7.I1.26; 7.30.22; 7.32.I, 4, 25, 28, 29, 31. The reference to Peter of Alexandria at 7 .32.3 I, for example, can hardly be an addition made at the same time as 9.6.2 (the correct chronological place for the notice), since it contains more information than the later note. Its appearance does, however, make perfect sense if Eusebius wrote books seven to nine as a single block (though most of the current book eight is a later addition, including the reference to Peter in 8.I3.7), commenting on Peter as he came up in the narrative, directly or through association. So numerous and so integral to the content and structure of the history are all these passages that if we were to accept them as later additions, the only possible hypothesis would be that they were included as part of a complete rewriting of almost the entire work for the edition of 3 I 3/4, not as part of a simple revision to keep the work up to date. Yet apart from the replacement of book eight there is no evidence for alteration or revision on this scale in later editions. Like book six, book seven could hardly have existed as it does now simply with the later references removed, 'ending almost exactly where the first edition of the Chronicle ended' with a reference to the death of Aurelian, the accession of Probus (7.30.22), a section on two recent bishops of Laodicea (32.5-2I), and 'a brief statement of the names of the bishops who occupied the principal sees at the time of writing'. 34 Those who argue for a date of c.303 must also explain why Eusebius would have concluded his history with the beginning of the persecution in March/April 303, a perverse and ignoble conclusion for such a work (unless he happened to finish it in January or February!). It should also be noted that Eusebius' list of emperors at 7.30.22 omits Tacitus and Florianus, short-lived Augusti of 275-76, incorporating their regnal year into Aurelian's total. It appears to be a deliberate simplification and as such is understandable in 3I3, almost forty years later, but is rather harder to explain in the early 290s or even in 303. An early date of composition can only be maintained by positing massive revision at a later date-especially to books six and seven-revision for which there is no evidence. The same problems exist for an edition concluding with book eight. 35 CHRONIC! CANONES/HISTORIA ECCLESIASTICA

Barnes, 'Editions', p. 200, and C and E, pp. 129 and 145-46. The view of, for instance, Schwartz (cit. n. 12), pp. lv-lvi, and H. J. Lawlor in Hugh Jackson Lawlor and John Ernest Leonard Oulton (trans.), Eusebius, 34

35

Bishop of Caesarea. The Ecclesiastical History and the Martyrs of Palestine 2 (London, 1928), pp. 3-6. This hypothesis rests almost exclusively on an inconclusive comparison of the wording between the preface (1.1.2) and passages of books seven and eight (7.32.32, 8.1.1, 16.1).

I 486 No one has produced any evidence that any passage outside of book eight is in fact a later addition, apart from the hypothesis of an early edition itself, which is simply petitio principii, arising out of the need to dispense with evidence contradicting an early edition. This is, I believe, the Achilles heel of any argument that places the composition of the HE before 3 I 3: there is no evidence that Eusebius ever carried out massive revisions of the sort required for the argument of an edition earlier than 3I3/4 and the burden of proof must lie with those who claim that he did. I can find no solid, objective evidence to suggest an edition earlier than 3 I 3, which would therefore be identical to Barnes' 'second edition': books one to seven, the preface to eight with the short recension of the Martyrs of Palestine, Galerius' edict (the 'palinode'), the 'appendix', and book nine. 36 If this is so, and the first edition of the HE was written in 3 I 3/4, then there is nothing to contradict the dating of the Canones proposed above and its first edition must therefore date between 306 and 3 I 3/4·

IV The reference to the Canones in the preface to book one of the Eclogae propheticae, that is to book six of the General Elementary Introduction, provides a general confirmation of this date for the Canones. This work could date to any period of general persecution between March/April 303 and May 3 I I, and December 311 and May 3 I 3, since it refers to the suppression of Christian worship and the detention of bishops, though Eusebius was imprisoned for a time during the second bout of persecution, making this later period less likely. As noted above (n. I), book ten of this work would seem to date after 309. Whatever the date, Eusebius' careful explanation of his methodology in composing and arranging the Canones implies that he was still in the process of completing the work and that the reader would have to take his word for it that he had proved the antiquity of Moses and the succeeding prophets. 37 It seems obvious from this unique descriptive reference that Eusebius did not expect his readers to be familiar with the work; it is quite different from his reference to the Canones at the beginning of the HE, where he assumes that his audience is familiar with it, just three or four years later. We unfortunately 36 Barnes, 'Editions', p. 201. This is the conclusion of J. B. Lightfoot and B. F. Westcott s.v. 'Eusebius', in William Smith and Henry Wace (eds.), A Dictionary of Christian Biography, Literatures, Sects and Doctrines 2 (London, r88o), 322-23 (which is accepted in principle by Lawlor (cit. n. 35), p. 3), Lane Fox (cit. n. 3), p. 6o8, and Louth, 'Date', p. 123, among others. 37 He starts off with 'Let it be known .. .' ('la·riov).

I CHRONIC! CANONES/HISTORIA ECCLESIASTICA 487

cannot tell how much further than Moses he had got when he was writing this part of the General Elementary Introduction. Since we cannot date the General Elementary Introduction with any great precision, it cannot help with a specific date for the Canones, but the two works do appear to have been contemporary, which is an important confirmation of the general date of the Canones put forward here.

v Eusebius' stress on the Years of Persecution, which start from the beginning of the persecution in Caesarea in March/April 303, and their use as the basic chronological system in both the Canones and the Martyrs of Palestine, in contrast to their very infrequent appearance in theHE(7.32.31 (ninth year); 8.13.10 (second year); 16.1 (tenth and eighth years)), 38 suggests a close correlation in conception between the first two works. Eusebius' expedient of doubling up the second and third Years of Persecution, rather than just omitting the system altogether and avoiding the entire problem, suggests that this section was written at a time when he believed that a close accounting of the individual years of the persecution was of great importance. The evidence of the different uses of this system in the existing book eight of the HE (it does not appear in book nine) and the two recensions of the Martyrs (on which, see Appendix I) indicates that this would have been before 313/4, probably at the same time as the Martyrs was being completed. If the Canones and the Martyrs are seen as complementary, it would help to explain why Eusebius makes no mention of events during the persecution in the Canones apart from the deaths and accessions of emperors. 39 This really only makes sense if he was relying (or expecting to rely) on another narrative (the Martyrs) to provide the details. One can hardly imagine Eusebius' concluding the Canones at a date before the Martyrs and not commenting in some way on the persecution he saw around him. The HE, on the other hand, presupposes both works and is an advance on both: a full narrative of Christian events, instead of an epitome, with a revised chronology, combined with an epitomized and rewritten version of the Martyrs as one of its chapters.

38 The first two of these resemble entries in the Canones, 227k and zz8d (the martyrdom of Peter of Alexandria and the abdication of Diocletian and Maximian). 39 For this, see my paper cited in n. 6, above.

I 488

VI A date for the Canones between 3o6 and 3 I 3 is further confirmed by Eusebius' own comments on its genesis and by its strong apologetic nature. In his preface to the Canones (reproduced and translated in Appendix 2, below) Eusebius makes it clear that it was Porphyry of Tyre's contradiction of established Christian and Jewish chronologies regarding Mwa ws apxatOTYJS in book four of his KaTii Xpwnavwv (Against the Christians) that led him to undertake his own chronological researches in the first place. 40 Eusebius' researches revealed that both were incorrect, and he drives the point home again and again throughout his preface. The question of Moses also appears in the preface to the Chronographia, where it is listed as the first goal of his work (Arm. r. I 6- I 8 = Grk. I67.I8-2o (seen. 2)). The importance of Porphyry's chronology for the date of Moses is emphasized again in Eusebius' discussion in the Praeparatio E Ka! 'IovaTo>, loCw> KaaTo> Ti]v &.7To8Etgtv lK 1raAau'is {moaxwv !aTop{a>. "Ivaxo> 8 Twv '!AtaKwv hww t1TTaKoa{ot> Louth, 'Date', p. u6. Lane Fox (cit. n. 3), p. 6o8, suggests that it was 'already prepared, but perhaps not circulated'. This, of course, left the Canones without any companion text to fill in the blanks of the persecution (see above). 80 Barnes, C and E, pp. 148-49. 78

79

I 504

o

11"pw{3Ev t. 'EA.IrqvtKwv OE cfnAoa6tPwv, 5ans 71"0TE EKI'Lvos &viJp TTJV Ka8' UVUK VTJV 11"po{3 {3AT]f-!EVOS, lv Tf) li' TijS E1S p.riTTJV alJT0 11"0VTJ(JdaT]S Ka(J' {mo8iaEWS 11"p0 TWV :E€p.tprip.EWS xp6vwv TOV Mwvaia yEvia8at tPTJa{· {3aatAEV t liE ltaavp{wv :EEp.{pap.ts 11"p6a8Ev v' 11"pos Tols p'. waT Elvat KaTa Toihov Twv TpwtKwv Mwvaia 11"pEa{3vnpov v' Kat w' 'Eyw liE 11"Ept 1foAAov Tov &ATJ8fJ A6yov np.wp.Evos Kat To 6.Kpt{3Es &vtxvEvaat lita U1fOVlif)s 11"poMJip.TJV. opp.TJ8 tS EV f-!EV Tf) 11"p0 TaVTTJS avvTag t Mas p.avT0 xpovwv &vaypatPas avvEAEgrip.TJV 1faVTo{as, {3aatAdas T€ Xa/..lia{wv, ltaavp{wv, M'Tjowv, liEpawv, Avliwv, 'E{3pa{wv, Alyv11"T{wv, JtOTJva{wv, ltpydwv, :EtKvwv{wv, AaKEliatp.ov{wv, Kopw8{wv, f?hTTaAwv, MaKEOovwv, AaT{vwv, o!s VUTEpov yiyovEV E1f{KATJV ovop.a 'Pwp.aLO!. op.ov y{vovTa! ! 1• 'Ev OE T0 11"ap6vn E1ft TO abTo TOVS xp6vovs avvayaywv Kat &vn11"apa8EtS EK 11"apaAA7}Aov TOV 11"ap' EKrianp TWV hwv &pt8p.ov xpovtxov Kav6vos avvTagw E1l"OtT]arip.T]v ...

Syncellus, Ecloga Chronographia (Mosshammer, 73.1 1-74·3). Moses, a man of the Hebrew race, was the first of all the prophets to hand down in writing the oracles and divine prophecies about our Saviour, I mean Christ, and about the nations' knowledge of God that came about through Him. Men distinguished for their teaching, such as Clement, Africanus, and Tatian among Christians and Josephus and Justus among the Jews, have said that Moses flourished in the time of lnachus, each in his own way furnishing proof from ancient history. Now lnachus preceded the Trojan War by 700 years. But of the pagan philosophers, whoever that man was who put forth that written attack against us asserts in the fourth book of that work that he fruitlessly laboured upon against us that Moses existed before the time of Semiramis. Now Semiramis ruled the Assyrians 150 years before lnachus. According to him, therefore, Moses predates the Trojan war by 8so years. Now since I consider historical truthfulness and accuracy to be matters of great importance I proposed to investigate this matter with great effort. This was my starting point. In the first volume of this work by furnishing myself with the raw material [necessary for such a study] I gathered together all sorts of chronological records and the kingdoms of the Chaldaeans, Assyrians, Medes, Persians, Lydians, Hebrews, Egyptians, Athenians, Argives, Sicyonians, Spartans, Corinthians, Thessalians, Macedonians, and Latins (who were later called the Romans)-fifteen altogether. In this volume I have collected these chronologies in the same place and contrasted in parallel columns the numbers of years, [which I have placed] beside each nation. In so doing I compiled a chronological table ...

II

A CHRONOLOGICAL PROLEGOMENON TO RECONSTRUCTING EUSEBIUS' CHRONIC! CANONES: THE EVIDENCE OF PS-DIONYSIUS (THE ZUQNIN CHRONICLE)

T

en years ago, when I first conceived the idea of comparing Jerome's Latin translation to a reconstruction of Eusebius' original Greek version of his Chronici canones in order to determine what changes Jerome had made, it seemed to me a fairly easy task. Although the Greek original of Eusebius' chronicle no longer exists, we have Jerome, a reasonably complete Armenian translation, two Syriac epitomes, and many different Greek witnesses. There might be a few difficult decisions here and there involving the evidence of a single witness, but I felt confident that for the final section of the chronicle the result would be clear and straightforward. Those who have seen the result, which was published in 1999/ will know that the result was neither clear nor straightforward. As my Greek text enters the fourth century, it erupts into a forest of brackets and question marks, indicating where I was uncertain of the wording or the chronology. This problem arose chiefly because it turned out that Jerome had altered the text as he translated it to a degree hitherto unsuspected and because other witnesses to Eusebius, who had

earlier been content with copying his text alone, began to use other sources once Constantine entered into the narrative. Cursory examinations of earlier sections of the text demonstrated other obvious places where Jerome had altered Eusebius' chronology. For instance, he shifted Eusebius' date of the crucifixion by one year and moved the year of Cleopatra's death back four years, to correspond to the ab urbe condita date he had from his Latin source, all the while retaining Eusebius' regnal year chronology for the Alexandrians. As a result, Cleopatra dies in her eighteenth regnal year, yet her reign continues on without her down to year twenty-two. In spite of my difficulties in reconstructing the Canones in 1999, I still harbour a desire to attempt a reconstruction ofEusebius in Greek, in spite ofthe difficulties. I have undertaken some preliminary, mostly chronological studies in this direction, and I would like to take this opportunity to present some of my findings, though it must be recognized that my appearance at this conference is a sham since I can no longer even read names in Syriac, which was the limited extent of my knowledge back in 1999. Then I relied upon Sebastian

II A Chronological Prolegomenon to Reconstructing Eusebius' Chronici canones

one was faced with descending strings of regnal years between which were suspended extremely brief comments concerning historical events. As far as we know, no Hellenistic or Roman Olympiad chronicle had attempted to record every single year over thousands of years in this way. Eusebius thus lies at the confluence of two independent types of Greek chronography, secular Hellenistic olympiad chronicles and Hellenistic Jewish and later Christian apologetic chronography, a topic that I discuss briefly elsewhere/ and which I shall develop further in the preface to a book on Latin chronicles that I am now working on with Michael Kulikowski. Yet, however revolutionary Eusebius' efforts were, the fact is that his influence on Greek historiography was severely limited. The Chronographia was the last apologetic chronological compilation; the Canones was the last olympiad chronicle. Diodorus, Panodorus, and Annianus produced chronicles that were in essence reworked versions of Eusebius, attempts to bring his decidedly anti-millenarian chronology more into line with the standard view that placed the birth of Christ around the year 5500 since the creation of the world, but we know little about their form or content. The Chronicon Paschale, composed in 630, was the last work to contain olympiad chronology, but it was fundamentally just translated, augmented, and extended Latin consularia, relying on the recounting of the annual consuls for its chronological backbone rather than the descending regnal years of kings or emperors. It would be another 200 years before Theophanes produced something similar, but even he collected all his chronological markers for emperors in Constantinople, kings of Persia, popes, and bishops into a

Brock, Marina Greatrex, and chiefly Witold Witakowski for Syriac help. Here I must admit that I have relied completely on the Latin translation of the Syriac and the German translation of the Armenian. Although today we look upon Eusebius' two-part chronicle, the Chronographia and the Chronici canones, as a ground-breaking and novel work, the more I study his antecedents the more I am of the opinion that to contemporaries much of it would have seemed very familiar. The first part, the Chronographia, was no different in form from the apologetic chronologies of Julius Africanus, who in turn had developed his work from such earlier apologists as Theophilus and Clement; what set Eusebius apart from these earlier writers was that he made almost no original contribution, but simply copied everything from earlier authoritative texts. The secular content of the Chronici canones was similarly cribbed from earlier chronicles and epitomes, and to the contemporary reader would have seemed just as much a pastiche of earlier work as the first part. On the other hand, a number of aspects of Eusebius' work would have seemed distinctly odd to an early-fourth-century reader. The first would have been the juxtaposition of two such disparate works, one in the form of Christian apologetic chronography, the second to all intents and purposes a traditional Olympiad chronicle. The second would have been the presence of Biblical history within this Olympiad chronicle, narrated alongside what for us is Greek mythology. And finally, perhaps most revolutionary of all, was the form of that Olympiad chronicle. Instead of the normal structure of paragraph blocks with the chronological data written in lemmata or rubrics above, 30

II A Chronological Prolegomenon to Reconstructing Eusebius' Chronici canones

Chronicle of 724 and the chronicle of PsDionysius or the Zuqnin Chronicle. Unfortunately, nothing is to be seen of Eusebius' chronological structure in these two works, completely eliminated in the former and reduced to date lemmata in the latter. As was the case in Greek, Syriac chronicles became more annalistic breviaria of history than true chronicles like Eusebius'. And by that I mean that the annual accounting of events was abandoned, narratives became longer and more involved, and the chronology was reduced to short headings or incorporated into the syntax of the historical entries. For Eusebius being able to see every single year on the page and to see the synchronism among those regnal years was just as important as the text. One of the most important cruces in the long study of Eusebius' Canones has been the actual format of his text. Excerptors and epitomators give us no clue regarding this and so we must tum to the translators. Here we meet a problem because there are two completely different formats. The earliest is that of Jerome's translation. Here Olympiads and the individually marked decades since the birth of Abraham run down the left-hand side of each page or double-page spread, while the regnal years of the various kings, the so-called.fila regnorum, are set up on the left and right side of each page or double page spread, where up to nine separate kingdoms have their regnal years recorded, until at the end the sole remaining kingdom, Rome, takes up its position on the left. This leaves an open space, the so-called spatium historicum, in the centre of the page for the inclusion of historical events. 3 The second possibility is the structure of the Armenian translation and of the Syriac exemplar continued by Jacob of Edessa.

single rubric heading each year. He also relied heavily on narrative sources, which meant that much of his work abandoned the brief notes that Eusebius provided even for the most recent events. Most of the works now often called chronicles, such as the work of John Malalas, are not chronicles in the same sense as the Chronici canones at all, but epitomes, breviaria, and annalistic compendia. And not only was Eusebius' influence limited, it seems likely that an intact uncontaminated manuscript copy failed to survive even the fourth century. In this light we can only describe Eusebius' great experiment as a dismal failure in Greek. In the Latin West, however, it was a different story, and it was here that Eusebius enjoyed his most lasting legacy. Through the translation made by Jerome in 380-381 the olympiad chronicle was brought to the West and spawned many continuators. Jerome's format was merged with the native consularia genre and the result was the birth of an historiographical form that was not only to survive the fall of the empire but to go on to become the standard historical genre of the Middle Ages as well. In Syriac, we have a situation that lies half way between the Greek and the Latin nachleben. Eusebius had only one true follower (at least, that we know of), Jacob of Edessa, a chronicler who followed Eusebius' format exactly, without knowing that the Syriac translation that he was working from was a much altered reworking, not the original at all. He thought he was correcting Eusebius' errors but he was merely correcting those of some unknown Syriac editor. As in the West, Syriac translations inspired many later chronicles, and in two cases epitomes of Eusebius' work served as the basis for continuations, in the so-called

31

II A Chronological Prolegomenon to Reconstructing Eusebius' Chronici canones

sion that the differences between the two could be accounted for by assuming that Eusebius originally filled the single spatium historicum with up to three columns of text that Jerome and later translators read in a different order, depending on whether they read across or down the columns. Errors in the dating of individual entries and the omission of certain entries could be explained by the position of the entries in this type of format. This view was accepted and amplified by Mosshammer, the only other person to have studied this aspect of the Canones.8 There is nothing inherently implausible with this interpretation, and there is existing evidence to support it: Jerome's translation includes a number of pages that preserve multiple columns (esp. 20, 23, 29, 31, 43, 46-51, 53, 57, 64, 65 [two columns] and 103 [three columns]), and in his preface he complains about the difficulty in figuring out the 'ordo legendi', the order in which one was supposed read the text (Helm 5.5). Furthermore, Greek manuscripts were often written in multiple columns, so it is not impossible that Eusebius would have written his text this way. The result is that both Helm and Mosshammer include sample pages of what Eusebius' manuscript pages would originally have looked like, though surprisingly both present pages written in minuscule, which was only invented 500 years after Eusebius' death. 9 The major problem with this multiplecolumn solution as I see it is that if Eusebius assigned certain dates to certain events and if the translators were attempting to copy those dates, as well as the content of the text itself, it shouldn't make any difference whether a translator read down a column or across the columns: the dates are the dates. It would only make a difference if the trans-

These run all the olympiads, years of Abraham (each one, not just by decade), and regnal years in that order down the centre of single pages and the spaces for the entries are added to the left and right of these multiple columns. There are no double-page spreads. 4 Because of the similarity of the formats and because of important linguistic indications in the Armenian translation that betray its Syriac origins, scholars have come to the conclusion that the existing Armenian translation of the Canones is probably the result of a collation between an earlier Armenian translation of a Greek text and a Syriac translation. 5 It used to be argued and accepted that the Armenian and Syriac structure was that of Eusebius' original Greek text. But following the detailed analyses of J. K. Fotheringham and Rudolf Helm, who both spent many years working on their editions of Jerome's translation, it became clear that it was Jerome's format that most closely mirrored Eusebius' origina1. 6 Today no one believes that Eusebius' Greek Canones looked like the Armenian or Jacob ofEdessa, and it is usually assumed that the transition from Eusebius' multiple fila regnorom to the central-column format took place in the earliest Syriac translation. However, Alden Mosshammer suggests that it occurred in an early Greek reworking of the Canones, perhaps produced in Alexandria at the beginning of the fifth century. 7 Much further confusion and consternation has been caused by the fact that Jerome and the Armenian translation do not always agree on the dates assigned to the historical entries and often do not agree even on the order in which these entries are listed. After much comparison of the Latin and Armenian texts, Rudolf Helm came to the conclu32

II A Chronological Prolegomenon to Reconstructing Eusebius' Chronici canones

text of Jerome, but six different manuscripts. My method was to compare the dates for every event that had a parallel in Jerome, Ps-Dionysius, and the Armenian translation. Only events dated in all three texts were considered. I also discounted any event in Ps-Dionysius that was not dated to a stated year or did not begin with the statement 'in the same year' or 'in this year'. Entries prefaced with statements such as 'at this time' were not considered since these entries almost always vary by ten or more years from the most closely cited date. The first conclusion of this comparison was that Jerome, Ps-Dionysius, and the Armenian almost always agree in their chronology when the event is a royal death or accession that is described within the fila regnorum themselves. There are some instances of a lack of agreement but these can be explained by scribal errors or other obvious modifications. Since the fila are continuous strings of numbers that are only broken by the accession of a new king, this is what we would expect. Each event is pegged to a specific year within the string and cannot move as long as the translator or copyist did not change the actual chronology. As a result I have not included any of these events in my analysis; I have included only those entries that appear within the spatium historicum and do not relate to accessions, deaths, or other events that are pegged specifically to the fila regnorum. This exception unfortunately reduces the total number of entries for analysis to a meagre 210. A comparison of these 210 dated events found that sometimes all three texts agreed with one another, while at other times all disagreed with one another. Sometimes two texts agreed but not the other. Nor was there a pattern with regard to the concentration of

lators didn't care where they stuck their entries: reading across three entries assigned to the same year and then writing them into three consecutive years for instance. But if that were the case, then it would not matter whether there were columns or not: the differences could be accounted for by copyists who did not care where they wrote their entries. The second problem is that there are also many discrepancies between Jerome and the Armenian translation in places where there are few entries, where there could never have been multiple columns to cause confusion. Now I do not doubt that Eusebius wrote some entries side by side, or in columns, or placed certain types of entries off to the side as a way of highlighting them (which Jerome then missed), but it seems to me that the theory of multiple columns cannot solve the problem of the serious discrepancies between Jerome and the Armenian translation. To come to a definitive conclusion I subjected this problem to careful analysis and in order to provide a control against the dates found in Jerome and the Armenian translation I included in my analysis the chronicle of Ps-Dionysius, which begins with an epitome ofEusebius' Canones. What marks out the epitome of Ps-Dionysius from the other more well-known Syriac epitome in the Chronicle of 724, the only Syriac text that Helm considered in his edition of Jerome, is that it provides dates for almost all its excerpts from Eusebius, using years of Abraham. In this I am, in a sense, following von Gutschrnid's 1886 comparison of the chronologies of these three texts, but he had different goals and used a different method of analysis. 10 He was also hampered by the fact that he had no reliable edition of any of the texts and for comparison used not a single 33

II A Chronological Prolegomenon to Reconstructing Eusebius' Chronici canones

the entries on a page; there could be agreement or disagreement regardless of whether the page was filled with entries or contained only a few scattered entries, and there was no pattern that suggested reading errors caused by multiple columns. Of the 210 shared spatium historicum entries among Jerome, Ps-Dionysius, and the Armenian translation, only 30 entries, or 14.3%, are dated to the same year in all three texts (see Table 1A). Jerome agrees with PsDionysius against the Armenian 43 times, which indicates that the Armenian is incorrect 20.5% of the time; Jerome agrees with the Armenian against Ps-Dionysius 40 times, which indicates that Ps-Dionysius is incorrect 19.0% of the time (about the same as the Armenian); and Ps-Dionysius agrees with the Armenian against Jerome 29 times, which means that Jerome is incorrect 13.8% of the time. Sixty-eight entries, or 32.4% of the total, have different dates in all three texts. Now if we shift the criterion from an exact match of dates to one of allowing a one year difference either way among the three texts, so that, for instance, Jerome could assign an event to 1345 Abr., PsDionysius to 1347, and the Armenian to 1346 (the assumed correct date being 1345), or Jerome and Ps-Dionysius to 1345 and the Armenian to 1346, the number of completely different dates drops from 68 to 14, or 6.7%, and the number of triple agreements rises from 30 to 122 (or 58.1%) (see Table 1B). Jerome agrees with Ps-Dionysius (i.e. the Armenian is incorrect by more than a single year) 19 times or 9% and with the Armenian (i.e. Ps-Dionysius is incorrect by more than a single year) 37 times or 17.6%, and Ps-Dionysius agrees with the Armenian (i.e. Jerome is incorrect by more than a single year) 18 times or 8.6%, which is statisti-

cally the same as the Armenian's error rate. The differences between A and B in the table show that while the mistakes that appear in Jerome and Armenian tend to be within a year, those in Ps-Dionysius tend to be larger than one year: the relaxation of the criterion produces a drop of only three entries for PsDionysius, yet 24 for the Armenian and 11 for Jerome. The remaining level of error is much higher for Ps-Dionysius as well, about double that of the other two witnesses. These patterns suggest something that no scholar has so far suggested and it has nothing to do with columns or reading up or down. In a normal olympiad chronicle the page is graphically divided by the chronological notice that presents the olympiad, the archons, the consuls, or any other chronological system. The entry for that year then follows in a single block. No matter how the text is copied the chronology can never change. Although Eusebius' text was a great step forward with regard to its presentation of a great amount of information in a small space, especially the graphic inclusion of every single regnal year for over 2,300 years, the spatium historicum was not graphically divided in any way and there was nothing to tie any entry to any particular regnal year on the left or right other than its position on the page (and perhaps the impressions of the ruled horizonatal lines used for writing). Since there was no way of anchoring the historical entries to the regnal years in a graphic manner, from the moment Eusebius' chronicle was first copied entries could start floating on the page, drifting up or down from one regnal year to another. This is particularly true on pages where the regnal years appear on every line of the text. Even the smallest slip of a single line could 34

II A Chronological Prolegomenon to Reconstructing Eusebius' Chronici canones

was the text that shifted. Furthermore, scribes might not even have realized that Eusebius intended the first line of each entry to be opposite a specific number in the fila regnorum. For such a scribe, it may have been that as long as an entry in his copy was in the same general area it appeared in his exemplar he was happy. Even the great German scholar Eduard Schwartz denied that the fila regnorum were intended to be read in any more than a general fashion and refused to accept that Eusebius would have tied the entries to specific regnal years. 11 In addition, since it seems obvious that the fila regnorum were written out first and the entries added later, any scribe who failed to maintain the correct spacing of letters and words within an entry could end up with entries too long or too short. Any closely following entries would then be dated too late or too early, as scribes paid more attention to the relationship of the text blocks on the page to one another than to their relationship to the regnal years. Since there could be over thirty regnal years per page, the range for error was therefore enormous. Jerome's complaint about the 'ordo legendi' is still valid: Eusebius no doubt wrote his entries in short lines or small text blocks all over the page. For one used to reading a text in neat lines within neat columns Eusebius' apparently haphazard text placement, combined with the fact that one had to read across double-page spreads for synchronisms and down to advance through time, must have confused all new readers of the text as much as it confused Jerome. Next in the process we must consider exactly the same sort of problems with regard to the translations themselves. Helm has a special apparatus to show opposite which regnal year each entry appears in the

cause the shift of a year. This became more of a problem on the double page spreads where the spatium historicum was quite wide and the entries quite short, and so the empty space between the edges of the written text and the regnal numbers was larger as well. As a result, every time the Greek original was copied every entry was subject to further potential shifting, up or down. This is probably what gave rise to one of the complications that Jerome mentioned regarding his Greek text of the Canones. He said that there were lines, or 'virgulae', all over the pages connecting entries (the 'res') to the regnal years (the 'numeri') (Helm 5.34). This obviously arose as different readers had compared one manuscript with another and used these lines to correct what they took to be entry creep in the manuscript that Jerome ended up using and its progenitors. The same phenomenon can be seen in some manuscripts of Jerome. In addition, most entries extend from their own year, indicated by the location of the first line of the entry, down through one or more following regnal years, and scribes could easily and mistakenly treat several different consecutive entries, each supposed to begin opposite a specific regnal year, as if they were in fact one large block of text dated to the year opposite the first line, failing to notice where individual sentences began within the block. This is very common in Latin manuscripts of chronicles, where individual entries were jammed together into a single text block to save space. As a result the marginal regnal years lost the obvious connection with the text opposite and begin to drift up or down the margins, erroneously taking the text opposite with them. In Eusebius' text it was the just the opposite: the regnal years were fixed and it 35

II A Chronological Prolegomenon to Reconstructing Eusebius' Chronici canones

many different Latin manuscripts. The same problem must have occurred with the Greek, Armenian, and Syriac manuscripts as well. There is a further problem with the Armenian translation in that it has two locations for the entries, one on either side of the column of regnal years. As a result entries could independently move up or down, thus seeming to alter the sequence of entries in relation to Jerome as well as their absolute chronology. It is also clear from my study in 1999 that crowding on one side or the other has shifted entries further down than they should be and in many cases caused the loss of entries that just didn't fit the narrow confmes left for them. 12 In this study I have also found that in some cases dense text blocks forced the upper entries into empty spaces above, thus ante-dating them. In view of these problems, it is really a testament to the care of the Greek scribes, the translators, and the later copyists that almost half of the surviving entries only vary from one another by a year, hardly a centimeter or two in the original Greek manuscripts. This study therefore demonstrates, first, that multiple columns are not necessary to explain the chronological differences in the translations. Second, it shows that there is no 'parallel corruption' shared between the Armenian and Ps-Dionysius, as one might expect since both derived from the same later redaction of the Canones. It is Jerome and the Armenian that share the same low level of error and Ps-Dionysius that is the odd one out. Whatever the nature of the common source of the Armenian and PsDionysius, its chronology was not modified from that of Eusebius' original in any way that is now evident. Third, it also suggests the solution to another larger problem.

It has always been assumed that the disappearance of Eusebius' original Greek text was due to Eusebius' anti-eschatological chronology, which set it apart from all other world chronologies of the time and made it the target of correction many times over the years. But that chronology could easily have been altered in a few places by simply changing the calculation figures or adding a supplement to account for the years between creation and Abraham, years omitted by Eusebius. His chronicle would not have to have been abandoned completely. This study, on the other hand, suggests another more obvious reason: it was just too complicated to be read easily and too complex a document to be copied accurately and economically. For later readers it made no sense to take up the space and the copying expense of recording regnal years in which nothing happened. As a result from a very early date I suspect that the early 'correctors' of Eusebius, like Diodorus, Annianus, and Panodorus, were not just changing his chronology from the creation of the world, as we know they did, but were also simplifying his fila regnorum into a more easily interpreted, more easily and cheaply copied, and less corruptible format. And so Mosshammer is almost certainly right, that the structure we see in the Syriac and Armenian witnesses goes back to an early Greek recension of the Canones and was not a Syriac innovation. But this takes us back to my original reason for undertaking this study. Given this obvious lack of agreement, how can one reconstruct Eusebius' Greek original? Obviously, it would be best to accept the date wherever two witnesses agree against the third, allowing for an error of one year. But that only gives us a date for 196 entries, out of hundreds and hundreds. And many puz-

36

II A Chronological Prolegomenon to Reconstructing Eusebius' Chronici canones

zling and unusual problems still remain. For instance, both Ps-Dionysius (1416 Abr.) and Syncellus (286.8-9, Mosshammer) record the foundation of Perinthus at the same time as Camerina, listing Camerina first. Perinthus therefore certainly appeared in Eusebius even though it does not appear in the Armenian translation (Camerina appears in 1417 Abr.). All manuscripts of Jerome mention the foundation of Camerina in the equivalent of601 BC (1416 Abr.) but, like the Armenian, most do not mention the foundation of Perinthus. However, four do. One puts it in the equivalent of 602 BC (1415 Abr.), which is where Helm puts it. But the three others put it in 60 1 BC, one including it within the same entry and after Camerina. The only conclusion can be that Jerome originally missed the entry but someone later compared his translation to a Greek version and added the entry in the margin. Later copyists put it in slightly different places. That it is missing in Jerome and the Armenian can only be a coincidence. There is other clear evidence of Jerome's text having been corrected against the Greek as well, though not in these manuscripts. 13

Another problem occurs in 735 BC (1282 Abr.) where both Ps-Dionysius and the Armenian translation locate the foundations of Syracuse and Catana in Sicily in a single entry. Jerome not only separates the entries, repeating the shared wording, but dates the first to 738 BC and the second to 736 BC, three years and one year earlier than the other two translations. The next entry, the capture of Messene by the Spartans, is dated to 735 BC (1282 Abr.) in all three texts. Unfortunately there is no Greek witness to help sort out the problem. Has Jerome spread these three entries out, moving them up into the empty space above, or did the Greek editor of the version behind PsDionysius and the Armenian compress them? So three steps ahead and two back. The result is that no accurate reconstruction of the Canones can ever be made, but if one accepts the evidence of Ps-Dionysius then a closer approximation can be produced. Whether that reconstruction ever will be made I cannot say, but until then it is certain that we can no longer blindly rely on just Jerome or the Armenian. The Syriac evidence must be given its proper due.

Table 1: Chronological Agreements Among Witnesses to the Chronici canones A. B. Exact Match One Year Allowance Three Agreements Jerome= Ps-Dionysius (=Armenian Incorrect) Jerome= Armenian (=Ps-Dionysius Incorrect) Ps-Dionysius =Armenian (=Jerome Incorrect) No Agreement Total

30 43 40 29 68

14.3% 20.5% 19.0% 13.8% 32.4%

210 100.0%

37

122 19 37 18 14

58.1% 9.0% 17.6% 8.6% 6.7%

210

100.0%

II A Chronological Prolegomenon to Reconstructing Eusebius' Chronici canones

NOTES canonum quae supersunt, vol. 2, ed. Alfred Schoene (Berlin, 1866), LIII-LV; Karst, Die Chronik, XXXVIII-LIV, and Alden A. Mosshammer, The Chronicle of Eusebius and Greek Chronographic Tradition (Lewisburg/ London, 1979), 50, 59-60, 73-5. 6 See the analysis of Mosshammer, The Chronicle, 38-73. 7 Mosshammer, The Chronicle, 75-9,80-1. 8 Rudolf Helm, Eusebius' Chronik und ihre Tabellenform, Abhandlungen der Priissischen Akademie der Wissenschaften. Jahrg. 1922. Philosophisch-historische Klasse, nr. 4 (Berlin, 1924), 1-56; Helm, Die Chronik, XXVIIXXXVIII; Mosshammer, The Chronicle, 62-4, 81, 82-3. 9 Helm, Die Chronik, XXX-XXXI and XXXVII, and Mosshammer, The Chronicle, 27. 10 Alfred von Gutschrnid, "Untersuchungen iiber die Epitome der Eusebischen Canones," in Kleine Schriften von Alfred von Gutschmid, vol. 1, ed. Franz Riihl (Leipzig, 1889), 483-529, at 492-524. 11 See Mosshammer, The Chronicle, 55. 12 See the commentary to the work cited in n. 1, pages 47-58, and my "The Dates and Editions of Eusebius' Chronici canones and Historia ecclesiastica," JTSNS 48 (1997) 473 n. 7. 13 See Mosshammer, The Chronicle, 52-3 for two famous examples.

1 Studies in Eusebian and Post-Eusebian Chronography, Historia Einzelschrift 135 (Stuttgart, 1999), 60-64. 2 "Apologetic and Chronography: The Antecedents of Julius Africanus," in Julius Africanus und die christliche Weltchronistik, Texte und Untersuchungen zur altchristlichen Literatur, ed. Martin Walraff(Berlin, 2006), 6-30. 3 See John Knight Fotheringham (ed.), Eusebii Pamphili Chronici Canones, Latine uertit, adauxit, ad sua tempora produxit S. Eusebius Hieronymus (London, 1923); Rudolf Helm (ed.), Die Chronik des Hieronymus. Hieronymi Chronicon 3, GCS, Eusebius Werke 7 (Berlin, 1984 [1956]), as well as my general introduction, "Jerome Explained: An Introduction to his Chronicle and a Guide to its Use," Ancient History Bulletin 16 (2002) 1-32. 4 See Josef Karst (ed. and trans.), Die Chronik aus dem Armenischen Ubersetzt mit textkritischem Commentar, GCS 20, Eusebius Werke 5 (Leipzig, 1911) and E. W. Brooks (ed. and trans.), Chronicon Jacobi Edesseni, in Chron. min. 3, CSCO 6, SS 6: SS 3.4, versio (Louvain, 1955 [1907]), 199-255; Syriac text: Chronicon Jacobi Edesseni, in Chron. min. 3, CSCO 5, SS 5: SS 3.4, textus (Louvain, 1955 [1905]), 261-327. 5 See H. Petermann's introduction to his Armenian translation in Eusebi Chronicorum

38

III JEROME EXPLAINED: AN INTRODUCfiON TO HIS CHRONICLE AND A GUIDE TO ITS USE

I

t

was probably in Antioch that Jerome first came across a copy of Eusebius of Caesarea's

XpovtKOl Kav6VES" ( Chronici canones or Chronological Tables) and he set about translating it

into Latin in 380 when he had arrived in Constantinople, adding extra material on Roman history and literature, and continuing it from its terminus in 325 down to 378. 1 His work of translation, augmentation, and continuation was completed before the middle of 381. His Latin translation proved to be more popular than the original Greek, for whereas Eusebius' original has perished almost without a trace, Jerome's translation survives in dozens of manuscripts and was used by countless Western writers in Late Antiquity and the Middle Ages for their histories. Being the only surviving complete and continuous history of the preClassical and Classical Mediterranean world it ended up being the foundation of all modem histories of those periods as well: Eusebius' Chronological Canons, translated from the original Greek into Latin by St. Jerome, became the basic chronological reference work in the Western world until it was superseded in the seventeenth century. Even then, with improvements by Joseph Scaliger and others, who integrated the various calendar systems into the Julian calendar, Eusebius' work in its broad outlines formed the basis of most historical writing on the ancient world2 Eusebius' Chronicle ...was translated into Latin, emended and extended by Jerome and became both a pre-eminent source and an influential model for all later efforts to reconstruct the chronology of the world. 3 It is one of the fundamental books upon which all research on the past of mankind has been based 4 It is doubtful if any other history has ever exercised an influence comparable to that which it has had upon the western world 5

Thus, from the late-fourth to the early-twenty-first century, our view of the past has been shaped by the way Eusebius and] erome originally described it. Even today the work remains an important historical source, as much for Greek history as for Roman. As Alden Mosshammer says, 'Sooner or later almost every classicist and historian of antiquity is confronted with evidence drawn from the Chronicle: of Eusebius.' 6

1 See ].N.D. Kelly, Jerome. His Life, Writings, and Controversies (London 1975) 72-5, and A.A. Mosshammer, The Chronicle ofEusebius and Greek Chronographic Tradition (Lewisburg PA 1979) 37-8, 67-73. For Jerome's Chronicle in general, see Alfred Schone, Die Weltchronik des Eusebius in ihrer BearbeitungdurchHieronymus (Berlin 1900), though in

many important respects it has been superseded. 2 D.J. Wilcox, The Measure of Times Past. Pre-Newtonian Chronologies and the RhetoricofRelative Time (Chicago 1987) 105. 3 A.

62. 4

5

Grafton, Joseph Scaliger. A Study in the History of Classical Scholarship, II. Historical Chronology (Oxford 1993)

J. Quasten, Patrology 3 (Utrecht/Antwerp 1960) 313. J.T. Shotwell, 'Christianity and History, Ill. Chronology

Psychology and Scientific Methods 17 (1920) 145. 6

(n. I) 30.

and Church History', Journal of Philosophy,

III 2

In spite of the importance of Jerome's Chronicle (as I shall call it here), the work remains little cited and less often consulted or read. Unfortunately the Chronicle is a late-fourth-century AD (and, worse, Christian) compilation, the product of a twilight world unknown to most Classicists. How is a Classicist whose idea of 'late' is Tacitus or Quintilian (or worse, Herodas) supposed to sort out such a foreign-seeming work? Those who are brave enough to seek out the Chronicle are often immediately put off by the complexity of the text and the multitude of arcane matters involved in understanding it, not least its chronology, one of its most important and valuable facets. It has been fairly described as 'one of the most labyrinthine and controversial texts in the field'? Given the importance of this history, the general ignorance amongst Classicists concerning it, and the errors that abound when it is used, it seemed to me that a description and explanation of the Chronicle and its manner of compilation might aid those who are unfamiliar with it but who must nevertheless take account of its evidence, whether literary or historical. It is hoped that this will make the use of the Chronicle a less formidable and perhaps even enjoyable task. I should like to begin with three examples that illustrate some of the problems scholars usually encounter with the Chronicle: Catullus, Sallust, and fourth-century AD ecclesiastical history, all three illustrating the problems of editions, chronology, and citation. Finally I wish to note the a priori assumption of error made by scholars when consulting the Chronicle and their hostility towards it. These will set the stage for the analysis and description that follow. l. The Problems l.l. Editions

When was Catullus born and when did he die? These are simple questions; the answers are very difficult. Since there is no explicit reference to the former (and certainly not the latter) in Catullus' own poetry, we must rely on external sources and allusions in Catullus' poetry for both. Of external sources, there is only one: Jerome's Chronicle. The majority of the works I have consulted that mention the problem of the dates state that Jerome dates Catullus' birth to 87BC and his death to 57,8 and most also mention that Jerome says that he was thirty years 7 Mosshammer ( n.1) 30. Note also T.D. Barnes, Constantine and Eusebius (Cambridge MA 1981) lll: 'Eusebius' Chronicle is not only a historical text of great importance for a wide variety of fields but also a literary text which poses problems of peculiar intricacy on several levels.' 8 The following is a selection in reverse chronological order of some of the more recent scholars who believe that Jerome dates Catullus' death to 57 BC: G.B. Conte, Latin Literature. A History (Baltimore 1994) l4 2; G. Lee, The Poems ofCatullus (Oxford 1990) xviii;J. Ferguson, Catullus (Oxford 1988) 10 (G&:R New Surveys in the Classics 20); H.P. Syndikus, Catull. Eine Interpretation (Darmstadt 1984) 1; G.P. Goold, Catullus (London 1983) 1-2; S.G.P. Small, Catullus. A Reader's Guide to the Poems (Lanham MD 1983) 1; M. Drury in E.J. Kenney and W.V. Clausen (eds. ), The Cambridge History of Classical Literature, II. Latin Literature (Cambridge 1982) 831; CJ. Fordyce, Oxford Classical Dictionary (Oxford 21970) 216; H. Bardon, Propositions sur Catulle (Brussels 1970) 5 (Collection La tomus 118);]. Michie and R. Rowland, The Poems ofCatullus (London 1969) 9; R.G.C. Levens, 'Catullus', in Fifty Years (and Twelve) of Classical Scholarship (Oxford 1968) 361; C.J. Fordyce, Catullus (Oxford 1961) ix;J.W. Duff, A Literary History of Rome (London 1960) 228; W. Kroll, C. Valerius Catullus (Stuttgart 1959) v; E. D'Arbela, Catullo. I Carmi (Milan 1957) 13; G. Laurenza, G. Valeri Catulli- Horati Flacci. CarminaSdecta (Firenze 1949) 6 n.1; 0. Mogavero,

III AN INTRODUCTION TO JEROME'S CHRONICLE AND A GUIDE TO ITS USE

3

old when he died. A quick check· of the best and most recent edition of the Chronicle (published first in 1913 and reprinted in 1956; seep. 9 below), shows that Jerome does date Catullus' birth to 87 BC but actually places his death in 58 BC, not 57. Jerome also says that Catullus was 'XXX aetatis anno', i.e. he was years old, as the dates indicate (87 58 BC). 9 The latter error, that someone in his thirtieth year is thirty years old, is common in the translation and interpretation of Latin. 10 The former error is more peculiar. Could it be a problem of editions? The 1923 edition also indicates 58 BC, as does the edition of 1866. Both will be detailed below. In the second edition of volume 27 of Migne's Patrologia Latina,11 the next most recently published edition, the entry is also placed under the equivalent of 58 BC, though the second part of the entry does carry over into 57BC (col. 4 28). This volume was published in 1866 and reprints an edition by Angelo Mai that was originally published in 1833, though in that edition the entry is completely under the equivalent of 57 BC. 12 The partial correctness of the PL text, therefore, is actually just a typographical error. However, I have never seen the Chronicle cited from this rare 1833 edition, so it is unlikely to be the source of the 58/57 confusion. It is only in the first edition of Vol. 27 of PL that we find unambiguous evidence for Catullus' birth in 57 BC in a cited edition (coll. Unfortunately PL 27 was first published in 1846 and its edition of the Chronicle is reprinted from a collected works ofJerome published in its second edition between 1766 and 1772. The first edition of this text of the Chronicle was originally published in 1740.n All these early Antologia dd Carmi di Catullo (Torino 1947) ix (claims he is quoting Schoene's 1866 edition (seep. 9); G. Lafaye, Catulle. Patsies (Paris 1932) v; and W.K. Kelly, The Poems of Catullus and Tibullus (London 1891) I. E.T. Merrill, Catullus (Cambridge MA 1893) xiv; F. Plessis, LaPoesiclatine (Paris 1909) 143; and A.A. Barrett, 'Catullus 52 and the Consulship ofVatinius', TAPhA 103 (1972) 23, state 57 but note that some manuscripts have 58. 9 Since Catullus mentions events after 58 BC Jerome's dates must be wrong, and so most scholars place his birth in 84 having first fixed his death in 54, a date after the latest reference in his poems, though some scholars suggest 82-52; see D.F.S. Thomson, Catullus. Edited with aTcxtual and Interpretative Commentary (Toronto 1997) 3-4 (Phoenix Supplementary Volume 34); M. Rambaud, 'Cesar et Catulle', BFLM 10 (1980) 38 and 49; and J. Granarolo, Catullc, cc vivant (Paris 1982) 19-22. Since Catullus was twenty-nine when he died, the date of his birth should be 83 or 81, respectively. In spite of this the entry in the third edition of the Oxford Classical Dictionary (Oxford 1996), by Hans Peter Syndikus, says, 'He was probably born in 84 BC or a little earlier, and probably died in 54 BC (p. 303). Whatever the dates, one immutable fact derives from Jerome and that is that Catullus was twenty-nine when he died. He could not have been born earlier than 84 BC. In general, see the important analysis of the problem by T.P. Wiseman, Clio's Cosmetics (Leicester 1979) 179-82. 10 Thomson (n. 9) 3 claims that 'XXX aetatis anno' could mean either 'twenty-nine years old' or 'thirty years old'. There is no such ambiguity in Jerome, here or elsewhere. 11 Few people realize that there are in fact two separate editions of the Chronicle in PL and, indeed, of the entire series of PL and PG (Patrologia Graeca ). The first edition of PL 27 appeared in 1846 and was replaced by a new edition in 1866, which reprinted the text used for the 1857 edition of PG 19. On 12 February 1868 a fire gutted the Ateliers catholiques, Migne's publishing house, and destroyed the lead prinringplates for PL andPG, among many others (see R.H. Bloch, God's Plagiarist (Chicago 1994) 104-12). Enough copies of the 1866 edition of volume 27 seem to have remained after the fire that it was not reprinted as most other volumes of PL were, starting in 1877. 12 Angelo Mai, Scriptorum vetcrum noua collectio eVaticanis codicibus cdita, VIII (Rome 1833) 365. Fotheringham described this as the 'editio omnium prauissima' (p. VII of the work cited on p. 9 below). l3 Sancti Eusebii Hieronymi Stridoncnsis presbyteri Opera... , ed Domenico Vallarsi and Scipione Maffei (Verona 1734-4 2; second edition, Venice 1766-72), coll. 603-4 of volume eight of the first edition. And to make matters

III 4

editions are based on manuscripts that are not even cited in the more recent editions because of their inferiority. The situation I am describing suggests one of two things. Either the arguments concerning Catullus' death currendy presented by most modern scholars date to a period before 1866 (i.e. no one actually bothers to look up the text of the Chronicle but just copies earlier citations), or when most people do seek out the Chronicle to cite it, they use an edition that is over 260 years old, either because that is the edition cited in earlier studies or because it is the only edition that can easily be found Neither is a creditable practice. 14 What selfrespecting classicist would use an edition of Aeschylus, Thucydides, Tibullus, Tacitus, ordare I say it?- Catullus that was originally published in 1740? No one. No one should be citing] erome from a 260-year-old edition either.15 As Vincent J. Rosivach has said, also with respect to Catullan scholars, '[W]ephilologists are more than occasionally an incestuous and lazy lot, too ready to rely on our predecessors' commentaries instead of checking their sources ourselves'. 16 1.2. Chronology Sallust was born in 86 BC and died in 35. Such is the consensus and so said no less a scholar than Sir Ronald Syme17 and the two latest editions of the OCD.18 But what is the basis for this? Jerome is our chief, but not our only, source. As Syme observes, jerome dates Sallust's death to 36 BC, saying that it was a 'quadriennium' before Actium (31 BC). The year is therefore clearly wrong. The date of Sallust's death is therefore shifted to 35, which is four

worse, Vallarsi's edition was simply a revision of Arnaldus Pontacus' Leiden edition of 1606! 14 The honour roll of those who correctly cite 58 BC is rather shorter than the list inn. 8, though most still do not cite the edition they are using: Thomson (n. 9) 3 (but does not give the actual AD equivalent);J.H. Gaisser, Catullus and his Renaissance Reada-s (Oxford 1993) 2; T.P. Wiseman, Catullus and his World. A Reappraisal (Cambridge 1985) 190; K. Quinn, Catullus. The Poans (London 1973) xxxviii n. 5; Rambaud (n. 9) 37; E. Bickel, Lehrbuch der Geschichte der Romischen Literatur (Heidelberg 21961) 491; M. Schanz and C. Hosius, Geschichte der Riimischen Litteratur I (Munich 41927) 293 (Handbuch der klassischen Altertumswissenschaft 7); R. Syme, 'Livy and Augustus', HCPh 64 (1959) 40; W. Eisenhut, Catull (Munich 1956) 221; H.J. Rose, A Handbook of Latin Literature (London 1954) 139; R.J.M. Lindsay, 'The Chronology of Catullus' Life', CPh 43 (1948) 42; andRE 7A2 (1948) col. 2355. V. Ciaffi, Il Mondo di Gaio Valerio Catullo e la sua poesia (Bologna 1987) 5; Granarolo (n. 9) 18; and A.L. Wheeler, Catullus and the Traditions of Ancient Poetry (Berkeley 1964) 88, cite both dates ('58 or 57'), Ciaffi accepting 58 but claiming that the manuscripts say 57, Granarolo accepting 57 (see alsop. 28) while admitting that the manuscripts say 58. 15 Another egregious example occurs in A.E. Samuel, Greek and Roman Chronology. Calendars and Years in Classical Antiquity (Munich 1972) 251 nn. 1, 2, and 252 n. 3 (Handbuch der klassischen Altertumswissenschaft 1. 7), who cites and bases his arguments upon the 1866 edition of Alfred Schoene at a point where Schoene's text is demonstrably wrong (the foundation of Rome). He only mentions the 1923 edition twice in asides, and Helm's 1956 standard edition, not at all (on these editions, seep. 9 below). Such an error by a chronographer is surprising. 16 V.J. Rosivach, 'Some Sources of Error in Catullan Commentaries', TAPhA 108 (1978) 216. The article makes the same points I do here, but with regard to Catullan commentaries. 17 R. Syme, Sallust (Berkeley and Los Angeles 1964) 13-14. 18 Second edition 946; third edition 1348.

III AN INTRODUCTION TO JEROME'S CHRONICLE AND A GUIDE TO ITS USE

5

years before 31. 19 But Syme and others have never bothered to look at Jerome's date for Actium (orrather the deaths of Cleopatra and Antony, which stand for it). It is placed in the eleventh year of 33 Be), even though Eusebius placed it in his fifteenth 29)?0 Since Actium is dated to 33 BC, Sallust's death in 36 BC is a 'quadriennium' before Actium, an inclusive 'quadriennium' (i.e. 36, 35, 34, and 33)?1 Since the item about Sallust's death's being a 'quadriennium' before Actium is obviously the only information Jerome had to date the event, Sallust must therefore have died in 34 BC. Sallust was therefore born in 86 and died in 34. The dates are also preserved in two other sources, both deriving from a source that Jerome also had (on this see below, p. 27). Like Jerome, the Descriptioconsulum (seen. 62) and the Chronicon Paschalc (seen. 34) both date Sallust's birth to 86 BC (pp. 224 and 347.ll). They disagree on his death, however. The former dates it to 38 BC (p. 225) and the latter to 39 (p. 359.10). It is more likely that 38 BC appeared in the common source, but we cannot be certain. Jerome, because of his dependence on Suetonius' de uiris illustribus (seep. 27, below), is rightly assumed to be the more accurate in this case. This highlights a further problem with ascertaining exactly what dates Jerome actually assigns to any of his entries. For instance, P. McGushin22 lays out the information for Sallust's birth and death as follows (p. 1): birth, 'ann. Abr. 1931 Ol. 173.3/4 C. 669 85 B.C and death, 'ann. Abr. 1981 Ol. 186.1/2 A.U.C. 719 35 B.C. This does not correspond with any of Syme's dates and it incorporates AUC dates that do not even appear in the Chronicle. Seven years later,J.T. Ramsey23 summed up the same information this way (p.1): birth, 'ann. Abr.1931 01.173.2/3 86 B.C, and death, 'ann.Abr.1981 01.185.4/186.1 36 B.C. The figures start off the same but they end up with completely different Olympiads and dates BC. Something must be wrong here and it seems to do with the relationship between 'ann. Abr.' (years from the birth of the patriarch Abraham; see below) and Olympiads. The answers to these and other puzzles will be provided below.

19 Syme also says that Jerome places Sallust's birth in 'the year corresponding to 87' (p.13), but this is just a typographical error for '86'. 20 This is because Jerome's source for Augustus appears to have dated the beginning of his reign to 722AUC (ab Urbe condita), which, subtracted from Jerome's date for thefoundation of Rome, 755 BC, provides a date of 33 BC, rwo years early. This source was the now-lost Kaisergeschichte (described below), as reported by rwo witnesses, Aurelius Victor: 'Anno urbis septingentesimo fere uicesimoque duobus etiam mas Romae incessit uni prorsus parendi' (1.1) and the anonymous Epitome de caesaribus: 'Anno urbis conditae septingentesimo uicesimo secunda, ab exactis uero regibus quadringentesimo octogesimoque, mas Romae repetitus uni prorsus parendi, pro rege imperatori uel sanctiori nomine Augusto appellato' (l.l). In spite of changing the date of the Battle of Actium,Jerome left Cleopatra's regnal years as they appeared in Eusebius. That is why she has four regnal years after her death! 21 On this point, see the highly sensible but almost entirely neglected article by G. Perl, 'Sallusts Todesjahr', Klio 48 (1967) 97-105, though even Perl fails to note Jerome's date for the Battle of Actium. 22 P. McGushin, C. Sallustius Crispus Bellum Catilinae. A Commentary (Leiden 1977). 23 J.T. Ramsey, Sallust's Bellum Catilinae (Chico CA 1984) 24 (APA Textbook Series 9).

III 6

1.3. Citation

And lest any reader think that I, as a historian, ·am criticizing just philologists for their sloppy practices, be assured that the same confusion holds true, unfortunately, for many historians who either ignore the text in favour of narrative histories or just copy out old references and cite the ancient text of PL. Even a work as recent as 1988 that deals with ecclesiastical history, for which Jerome's account is of fundamental importance, cites entries in the Chronicle in the following ways: '2337 years after Abraham', 'sub ann. 328' (this is the standard formula), 'sub ann. Constantius XII [?353] PL 27: and 'sub ann. AUC 2370, reign of Constantius XII (?353) and cites the tide three different ways: Continuation ofEusebius' Chronicle, Chronicle, and Chronicon. 24 These mixed up references show that the author has confused two completely different editions in his citations, that of PL and Rudolf Helm's 1956 edition, in the last example above citing the text of PL with Helm's page numbers. In the first example he shows that he knows that the fundamental dating system used in the chronicle is years since the birth of the Hebrew patriarch Abraham (as we shall see), but in the last he thinks that they are years ab Urbe condita (AUC 2370 is AD 1617!). He also states that '2337years after Abraham' is AD 320, when it is in fact 321. Three of the four examples use the form 'sub ann.', yet each refers to a different type of date: years AD, which are editorial additions and appear in the margins of only Helm's edition; regnal years; and 'Years of Abraham'. I could continue to cite such examples (a few more are noted below inn. 67), but I think that I have made my point. It is clear that where Jerome's Chronicle is concerned fear and confusion reign supreme, for philologists and historians alike. 1.4. The Assumption of Error Jerome compiled and translated the Chronicle very quickly. He admits that himself (seep. 29, below) and observable errors in his translation from the Greek confirm it?5 Historical analysis has revealed many apparent errors in his chronology as well?6 The result of these observations is that many of those who use the Chronicle feel compelled to make some kind of derogatory comment about it or Jerome. This is an attitude of long standing, and a latent (and not so latent) hostility towards the chronicler is often revealed as well: Nam sane si quis hodie ita, ut fecit Hieronymus, Graeca uerteret, non dico alienis, sed ut Actaeon, a suis canibus mordicus discerperetur (If someone translated Greek nowadays as Jerome did, he would be tom apart, not only by the teeth of others' dogs, but, like :Actaeon, by hisown)?7 ... his carelessness and haste are evident not only in his numerous errors (even of translation), but also in his apparent indifference to exact dating even where the material for it was 24 R.P.C. Hanson, The Search for the Christian Doctrine of God (Edinburgh 1988) 132, 277 n. 5, 341 n. 106, 382 n, 132, 399 n. 72, 461 n. 8, 761 n. 117, 793 n. 12. The author was a respected ecclesiastical historian. 25 On these see Grafton (n. 3) 575-7, and Mosshammer (n.l) 52-3. There are others. 26 The longest sustained investigation is Rudolf Helm, Hieronymus' Zuslitzdn Eusebius' Chronik und ihr Wert ftir die Literaturgeschichte (Leipzig 1929) (Philologus Supplementband 21, 2). For a detailed analysis of a short section from AD 284-325, see R.W. Burgess, Studies in Eusebian and Post-Eusebian Chronography.!: The Chronici canones of Eusebius ofCaesarea. Content and Chronology, AD. 282-325 (Stuttgart 1999) 90-8 (Historia Einzelschriften 135). 27 Joseph

Scaliger in 1658, quoted and translated by Grafton (n. 3) 577.

III AN INTRODUCfiON TO JEROME'S CHRONICLE AND A GUIDE TO ITS USE

7

provided by his sources and in his refusal to take the trouble to fill obvious gaps in these. Further, had he not been in such a hurry, or had he been interested, he could have made his chronology of Roman history more precise by incorporating references to the consuls. 28 Jerome, it is clear, operated in a casual and careless fashion. Where there are facts to check him, he can be convicted of gross errors... 29 Jerome's mistakes are notorious and in his treatment of Catullus he has made one blunder of a very elementary nature in the date of the poet's death. The rest of his information should, in view of this, be regarded with the utmost suspicion. 30

As we shall see, however, in spite of a few truly amazing blunders and mistakes, Jerome was far more accurate and competent than he has ever been given credit for, a conclusion that is revealed through an understanding of the Chronicle and its mysteries. We must now turn to the Chronicle itself in order to explain these mysteries. I shall begin with an account of its composition by Eusebius and then discuss the most recent and best editions of Jerome's translation and continuation. I shall then take the reader on a guided tour of the best edition available and explain its content, structure, and chronology, in the hope that it will become more familiar and understandable to the 'lay reader'. The final sections will deal more specifically with evaluating Eusebius' original Greek text and Jerome's additions to it. 2. Eusebius and the Greek Original May 339) was without doubt one of the most important Eusebius of Caesarea Greek scholars of the early church. His most influential and lasting contributions, to the eastern and the western parts of the empire alike, were in the field of history, for he both brought the Christian world chronicle to its final fruition and almost invented the genre of ecclesiastical history. His Canones was probably first composed around 311 as a work of apologetic and scholarship, aimed at Christians and alike. 31 The final and most influential edition was published in the latter half of 325 and then 28

Kelly (n. 1) 74-5. Syme (n. 14) 40. 30 Barrett (n. B) 29 n. 23. 31 For these points, see R.W. Burgess, 'The Date and Editions of Eusebius' Chronid canones and Historia ecclesiastica', ]TS ns 48 (1997) For the chronicle and the historiographical background to it, see Mosshammer (cit. n.1) 29-112 (an absolutely fundamental work with a full bibliography that anyone interested in Jerome's Chronicle must read); Heinrich Gelzer, Sextus Julius Africanus und die byzantinische Chronographie 1 (Leipzig 1880) and 2.1 (Leipzig 1885), 1-107; RE 6.1 (1907) 1370-1439, esp.1376-84; Quasten (cit. n. 4) 311-13;]. Sitinelli, Les vues historiques d'Eustbe de Ctsartc durant Ia ptriode prtnictrnne (Dakar 1961) 31-134; W. Adler, 'Eusebius' Chronicle and its Legacy' in H.W. Attridge and Gohei Hata (eds.), Eusebi115, Christianity, and Judaism (Leiden 1992) 467-91; R.M. Grant, Eusebius as Church Historian (Oxford1980) 3-9; B. Croke, 'The Origins of the Christian World Chronicle', in B. Croke and A.M. Emmett ( eds.), History and Historians in Late Antiquity (Sydney 1983) 116-31; idem, 'The Originality of Eusebius' Chronicle', AJPh 103 (1982) 195-200; idem, 'Byzantine Chronicle Writing 1: The early development of Byzantine chronicles', in E. Jeffreys with B. Croke and R. Scott (eds. ), Studies in ]ohnMalalas (Sydney 1990) 27-38; T.D. Barnes, Constantine and Eusebius (Cambridge MA 1981) lll-25; D.S. Wallace-Hadrill, Eusebius ofCaesarea (London 1960) 155-67; R.M. Grant, Eusebius as Church Historian (Oxford 1980) 3-9; W. Adler, Time ImmemoriaL Archaic History and Its Sources in Christian Chronography from Julius Africanus to George Syncdlus (Washington DC 1989) 15-71 (Dumbarton Oaks Studies 26); and Burgess (n. 26). 29

III 8

modified again slightly in the middle of 326. The Canones was actually the second of two linked works, the first being the Chronographia, a patchwork of narrative and regnal lists setting out nation by nation, as Eusebius himself said, the raw material for a complete chronology of world history from the time of the patriarch Abraham. In this work Eusebius sets out all his sources and the raw information that he derived from them, kingdom by kingdom, in the manner of the works of earlier Christian chronographers like Julius Africanus. The second volume, the Canones itself, is the synthesis and tabulation of the raw material in the Chronographia. It sets forth all known world (i.e. Mediterranean) history from the birth of Abraham, in what we would call20l6 BC, to AD 325, noting each regnal year of the kings of all the important Mediterranean kingdoms in parallel vertical columns first on double-page spreads, and then on single pages, to a maximum of nine kingdoms at a time. Important events and individuals from secular and Biblical history are noted under their proper regnal years or Olympiads, as well as Eusebius could calculate them on the basis of the often conflicting evidence that he possessed. Eventually all the columns resolved themselves into a single column of text representing the year by year chronology of the Roman Empire, the polytheistic polyarchy of the past resolving itself into the monotheistic monarchy of the reign of Constantine. Although Eusebius was writing within definite chronographic traditions, Hellenistic and Christian, no such universal synchronism for world history had ever been written before. Part of Eusebius' purpose in writing his two-volume chronological work was to oppose what had become a popular eschatological view that the world would last 6,000 years from Creation and that Christ had been born in the year 5,500 (half-way through the metaphorical 'sixth day' of Creation). Unfortunately, his revisionist views on this and other chronological matters were not accepted by most other historians and his new chronology came in for a great deal of criticism. A number of later authors reworked and modified Eusebius' chronology, notably Diodorus of Tarsus in the last quarter of the fourth century, Annianus and Panodorus at the beginning of the fifth century, Andronicus during the reign of Justinian, and Jacob of Edessa, a Syriac chronicler and writer, c. 692. The result of two centuries of intense chronological criticism and revision was that Eusebius' unaltered original became harder and harder to find. Eventually it disappeared altogether and today no longer exists. All that is left of the Canones is Jerome's Latin translation, a reworked Armenian translation in two manuscripts of the twelfth or thirteenth century that is a compilation of an earlier Armenian translation and a Syriac translation, 32 two s;::iac epitomes, 33 and excerpts from later Greek works that used Eusebius as a source. 3 The 32 Josef Karst (ed.), Die Chronik aus dem Armenischen ubersetzt mit textkritischem Commentar (Leipzig 1911) (Eusebius, Werke 5: Die griechischen christlichen Schriftsteller der ersten dreijahrhunderte 20). 33 J.-B. Chabot (trans.), Chronicon miscdlanrum ad annum Domini 724 pertinens (Louvain 1955 [1904]) 63-119 (Corpus Scriptorum Christianorum Orientalium 4, Chronica Minora 2: Scriptores Syri, series 3, tomus 4, versio), andJ.-B. Chabot (trans.), ChroniconPseudo-Dionysianum uulgo dictum (Louvain 1949) (CSCO 121: SS 3.1, versio). 34 Chiefly three: Ludwig Dindorf (ed.), Chronicon Paschale (Bonn 1832) (Corpus Scriptorum Historiae Byzantinae); A.A. Mosshammer (ed.), Georgii Syncdli Ecloga chronographica (Leipzig 1984); and A. Bauer (ed.), Anonymi Chronographia syntomos e codice Matritensi no.121 (nunc 4701) (Leipzig 1909).

III AN INTRODUCfiON TO JEROME'S CHRONICLE AND A GUIDE TO ITS USE

9

Chronographia has fared a httle better in this regard since large, continuous excerpts still exist in Greek and there is a complete and generally accurate Armenian translation as well. 35 Though most people have never heard of it, Eusebius' Chronici canones is one of the great losses from Greek and Roman antiquity. 3. The Editions There have been many editions of Jerome's translation, augmentation, and continuation of the Canones since Bonino Mombrizio's editio princeps of c.I475, but only three are of any value. I. Eusebi Chronicorum libri duo, edited by Alfred Schoene, Vol. I: Eusebi Chronicorum liber prior (Berhn I875) and Vol. 2: Eusebi Chronicorum canonum quaesupersunt (Berhn 1866). This is a gold mine of information for the serious scholar of the Chronicle but should be avoided by those unfamiliar with the intricacies of the manuscripts and different translations and traditions. Volume I is still the only comprehensive reconstruction of the Chronographia, though there is a more recent and more accurate German translation of the Armenian translation of this work (above, n. 32).

2. Eusebii Pamphili Chronici canones, Latine uertit, adauxit, ad sua tempora produxit S. Eusebius Hieronymus, edited by John Knight Fotheringham (London 1923). This was the first complete, printed edition, with text and apparatus, to represent the text as it was when Jerome composed it, with the spatium historicum and the twenty-six-hne page (on which, see below). These had appeared in Helm's 1913 edition, but in a hand-written form (in imitation of seventh-century uncials!) with no apparatus criticus. It can still be used with complete confidence for its text (in many matters of textual criticism it is better th,an Helm's edition) and it is still necessary for its complete apparatus criticus, though it is not normally cited for the text. 3. DieChronikdes Hieronymus. Hieronymi Chronicon, edited by Rudolf Helm (Eusebius, Werke 7: Die Griechischen christhchen Schriftsteller der ersten J ahrhunderte), third edition, Berhn 1984 (first edition: vol. I, GCS 24, Leipzig I913 [text], and vol. 2, GCS 34, Leipzig 1926 [apparatus]; second edition [typeset in a single volume]: GCS 47, Berhn 1956). This is the standard edition and normally the only one that should be cited, though only in the second or third editions. 36 It should be treated with respect, for it is a work of immense labour and intellect: the Chronicle of Jerome is the most difficult and comphcated text from classical antiquity ever edited, apart from the New Testament. That is why it took over 300 years to work out its structure and content.37 Jerome's complete Chronicle has never been translated into any modern language. A translation ofJerome's continuation from 326 has just been pubhshed with a commentarr8 and I 35 Karst ( n. 32); Alfred Schoene ( ed. ), Eusebi Chronicorum libri duo. Vol. 1: Eusebi Chronicorum liber prior (Berlin 1875); and, for the Greek, John Anthony Cramer (ed.), Anecdota Graeca e code!. manuscriptis bibliothecae regiae Parisiensis, 2 (Oxford 1839, repr. Hildesheim 1967) 118-63. 36 Actually the 'third edition' is just an uncorrected reprint of the second edition with an extra preface that provides an updated bibliography. 37 On this, see Mosshammer (n. 1) 38-83. 38 M.D. Donalson, A Translation of Jerome's Chronicon with Historical Commentary (Lewiston, Queenston, Lampeter 1996), a PhD dissertation published unaltered by the Edwin Mellen Press.

III 10

*

am myself working on a translation of, and commentary on, the text from 284 for the Translated Texts for Historians series. 4. Helm's Edition- An Overview This section is divided into three parts. The first part, 4.1, is designed to give all readers a general overview of the Chronicle as it appears in this edition. The next two parts, 4.2 and 4.3, are designed for those readers who are more interested in some of the details of the chronological structure of the Chronicle, and these can be skipped by general readers. 4.1. The General Tour Helm's edition of the Chronicle is divided into four main sections: the introduction, which includes the sigla; the edition itself; an index of names; and a collection of witnesses to Eusebius' text and other parallel sources for the whole text. The preface extends from p. IX top. XLVI and does what one would expect a preface to do: it describes the manuscripts and sets out the evidence for the structure and format adopted in the edition. It also includes a number of figures that give some idea of the original structure of the work in Greek (pp. XXX-XXXI, XXXVII), though thefont is rather small and the pages appear more empty than they would have originally. 39 Most of what Helm argues in his preface is accepted by the majority of modern scholars and so it is really an artefact from an earlier debate that he has now won. Next follows the list of manuscripts cited in the apparatus, with their dates. The most important manuscript here is 0, the Bodleian manuscript (MS Au ct. T. 2. 26), which was written within 75 years ofJerome's original. Next follow the various abbreviations used in the apparatus and the testimonia at the end of the book (pp. XLVIll to LII). The edition proper then begins on p.1 with Jerome's preface to his friends Vincentius and Gallienus. This extends top. 7 where Jerome's translation of Eusebius' preface begins. This continues to p. 19. In the top corner of each page, opposite the page numbers, are more numbers followed by an F, all in brackets. These continue throughout the text and are crossreferences to the equivalent pages in Fotheringham's edition (see p. 9 above, and Figures 1 and 2, below). These prefaces should be cited by page and line number (see the example below). Next comes the beginning of the chronicle itself. The first part, from p. 20 top. 105, consists of double-page spreads, indicated by Helm with a single page number for each spread and an appended 'a' for the left-hand page and a 'b' for the right-hand page. Thus the first spread is numbered 20a and 20b (I have included a reproduction of these two pages in Figure l and a later page, annotated, in Figure 2). Down each page are two columns of Roman numerals, making four in total for the spread. These are the fila regnorum, the regnal years of the kings and emperors of the important kingdoms of Mediterranean civilization. From left to right we have the kingdom of the Assyrians, the Hebrews, the Sicyonians, and the Egyptians, starting respectively with year 39 A better impression of an original chronicle page appears in the appendix to chapter two of R. W. Burgess, Studies in Eusebian and Post-Eusebian Chronography. 2: The Continuatio Antiochiensis Eusebii: A Chronicle of Antioch and the Roman Near East in the Reigns of Constantine and Constantius II, AD 325,350 (Stuttgart 1999) 174-7 (Historia Einzelschriften 135).

III AN INTRODUCTION TO JEROME'S CHRONICLE AND A GUIDE TO ITS USE

11

of Ninus, first king of all Asia except India; year one of Abraham the patriarch; of Europs, the second king of Sicyon; and the beginning of the sixteenth year dynasty of Egypt, during which the Thebans ruled Egypt (the actual pharaohs do not appear until the eighteenth dynasty, though they disappear again during the twentieth). The people or kingdom of each filum is labelled at the top of every new page so that the reader can keep track of them. Each new king, patriarch, high priest, pharaoh, dynasty, judge, or emperor is introduced with a rubric noting the kingdom, the sequence (second king, sixteenth dynasty, and so on, though this is not true for all kingdoms), the leader's name, and the length of his reign in years. This can be seen at the top of p. 20, but it can also be seen in its more usual form in the hand filum of the Assyrians after the year of Ninus, where years, succeeds Ninus. Semiramis, the second ruler of the Assyrians, who ruled for Individually these fila run from top to bottom, from one spread to the next. Moreover, all four fila line up in sequence across the pages to mark a single calendar year. On the side Helm notes in a smaller font the equivalent modern dates, starting with 2016 BC. Thus regnal year eight of Semiramis, eighteen of Abraham, of Europs, and eighteen of the sixteenth dynasty of Egypt are the rough equivalent of our year 1999 BC (on this, see pp. below). Here it can be seen as well that Helm follows the manuscripts exactly and uses the contemporary Roman numerals with 'Illl' and 'VIlli' for four and nine, for example, not the familiar 'IV' and 'IX' forms. These regnal years can be cited by modern scholars in a number of ways: 'Year 39 of Europs', '39 Europs' (the simplest), 'XXXVIIII Europs', 'Europs 39', or 'Europs XXXVIII!'. Abbreviations of names are often used when dealing with short chronological periods where the reader would easily understand which filum is being referred to: '39 Eur.'. On the far hand side, on p. 20a, to the left of the filum for Ninus, there is an underlined 'X', and all the Roman numerals across the page to the right are underlined as well. This 'X' marks the tenth year of Abraham (usually abbreviated as 'ann. Abr.lO', 'ab Abr.10', or '10 Abr. '), the standard relative chronology of the entire chronicle. From this point to the end of the chronicle every tenth year is underlined and the associated year from the birth of margin. The distinction helps to keep everything in line Abraham appears in the across the page. Since there was a wide diversity of dating systems in use in the classical world and none of them reached back to 2000 BC- the standard system of Olympiads went back only as far as 776 BC - Eusebius had to invent his own chronological system to hold the entire chronicle together. Some chroniclers in the Hellenistic tradition had counted backwards from the time of writing, like our BC system, but this did not appeal to Eusebius, and this is just as well, since each of his numerous updates would have necessitated a complete revision of the chronological accounting of the entire chronicle. Christian chroniclers often started their calculations with the Creation of the World and so used an annus mundi (AM) system that counted forwards from that point, but as was noted above Eusebius was trying to play down the need to calculate the age of the world (though he did include such totals in the Chronicle, as we shall see) and was also trying to avoid the complicated and contradictory chronologies of the early part of the Bible. As a result, he had to begin at some other point.

III 12

Fig. I

20a. Regnum .Assyriorum Primus omnia Asia.e exceptis India regna.vit Ninus, Beli filius,

ann -LII· huhJS ·XLIII· imperii anno natus est Abraham XLIII " Ninus condidit ciuitatem Ninum XLUII in regione .AssyriXLV orum, quam HeXLVI braei uocant NineXLVII uen XLVIII • Zoroaatres magus, XLVIlli rex Bactrianorum, L clarus habetur, adLI ,X LII uersus quem Ninus Assyriorum . II. dimicavit Semiramis ·ann ·XLII·

I

II III IIII

Hebraeorum gentis exordium In hilius imperio aput Hebra.eos nascitur Abraham, qui cum · C · esset o.nnorum, genuit Isaac Abraham • Regnante NiI no aput As· II syrioe nouisIII simo eius temIIII pore naecitur v Abraham VI 11 Abraham VII natione VIII Cbaldaeus VIIII primam aetatem aput Cbaldaeoe terit XI

XII XIII XI III XV XVl XVII XVIII

v

VI VII VIII

deest 0

(16 F)

• ad VI B, ad X AL

1 omnes B 2 regnat L 3 belli B 4 aii B anni A anniB M P ann. APN 6 abraam A HiM'. in 11,114 1 Polem. Silv. (CAr. m. I 641,11) Hrd(Cltr. m. II 13,19) CIJMiotl. 6 (Cltr. m. II 1110) 6 Citron. Gall. a. 611 (.CAr. m. I. 6311, 113) Yii:C • ..tquit. 7 (CAr. m. I. 681,16) CtJMiotlor. 6 (OAr. m. II 110) P1.-I1id. (OAr. m. II 498, 31) /lid. iun. 314 (Cltr. m. II 431) Aug.d.c.d.X VIII 2 (268,113 1170, 4) Oral. II 2,1: 3,1 lord. Rom.111[3 Beliehng dllr Konigllille zu der tDirJ:IWen (Mill. d. Dnllch. OriMJI.QusU.tclf. Nr. liB Augut 1917 B. SO) w nicltt vorlfanden 18 se1semiramis L an APM(L C111riod. 8 (CAr. m. II 120) Aug.d.c.d.X YI 17 (169,11.116) X VIII 2 (210, 9) I 4, 4 II 3, 1 YII 2,13 (43, 6; 86,14: 436, Hebr. 2 apud BMPN hebreos MPN 3 abrai L 4 centum anna18) lord. Rom.14 rum esset -M ann L 5 issaac .A 6 abrahae B abraam L s. Ass. 1-5 Ala.

o,_

a) t 7 condedit B 8 ninum (L 10 hebrei BM P 11 niniuen L ninniuen P •· O.n.10, 11 Ion.1, 2 Bier. in 0161 2,114 Clllliod. 7 (Oltr. m. II 1110) lord. Rom. Ill b) 7 ninus L 8 apud BM PN 9 88Birios LB 11/2 natus est atr L z. gropen Teilllef'Wiclten L 13 zoaatres LM e) 14 batrianorii A bactrianno.tr M 16 demicauit B obiidicabit L

10

l5

10

11

III AN INTRODUCfiON TO JEROME'S CHRONICLE AND A GUIDE TO ITS USE

20b

(17 F)

10

1&

110

13

Regnum Sicyoniorum

Begnum Aegyptiorum

In Graecio. vero ·II · Sicyonis imperabat Europs ·ann ·XLV· cui us regni ·XXII· anno natus est Abraham

Porro aput Aegyptios . XVI· potesta.s erat, quam vocant dynastiam. Quo tempore reg· nahant Thebaei annis CXC ·

XXII XXIII XXII II XXV XXVI XXVII XXVIII XXVII II XXX XXXI

• Nino regnante aput Assyrlos primos Sicyonis imperauit Aegia· leus annis ·loll·, a quo Aegialia nuncup&· ta eet, quae nunc Pe· loponneaus uocatur. Post quem secundus Europe, qui et prae· latus eat titulo

1

XXXII XXXIII

• Assyrlis imperauit Ull:or Nini Semiramis, de qua lnnumerabi· lia narrantur, quae et Asiam tenuit et prop· $er inundationem aggerea eonatrUll:it plurima Babyloniae urbis instaurana

xxxun

XXXV XXXVI

xxxvn

XXXVIII XXXVIlli

deeat 0

Nino regn&n· te aput Assy· rios Aegyptiis Thebs.ei impe· rabant

I II Ill III!

2018&.0

v

201ll

VI VII VIII VIlli

X XI XII XIH XIV XV XVI XVII XVIH

2007

1899

• ad XXXIII A B

Sic. 1 greeia LBM P

2 aycionis B aieioniis M imperauit (uit i. Rt18.) N· 4 ann N A.,. tl.e.tl. X VIII 2 (210, 4) Aeg. 1 apud BM PN aegiptioa B egyptios LM 2 XV• P XVII M erant L 3 dinastiam ABN 6 thebei PM tebei L anii BM an. L Ed. Me1/8f' 83

Hw. in 08u 2,24 Iaitlor. iun. 32 (Chr. m. II 431) Aug.tl.e. tl. XXI 14 (544, 18) Oroa. I 4, 3 d) 16 caldeus LP chaldeus AM B (u in o) 17 apud BM PN 18 caldeos LP chaldeos AM B (o OU& u) 19 tenet PN (verb. terit) gerit M lord. Rom. 14 e) z. gropen Teil verbliGAen L 7 regnante Nino -M 8 apud BM PN a 9 sycionis B aycioniis M BicyouJ N 10 imperabit L 11 ann. BM XLII N 12 aegilla N 13 peloponnens B peloponenaus L pelo15 II L A.,.tl.e.tl.XVII!B (ll10, 5) I) 7 ninus L 8 apud BMPN ponnisus M super L 9 aegiptios B 10 thebaeia B teuel L thebei M inperabant A g) t 19 lm· perabit L 20 ninis symiramis B 21 indmerabilia A 22 que B 24 inundationea L undationem M aggeria (a. Htl. verb.} A aquae aggerea B aggerea in aquae aggeres a. Htl. P 26 babilloniae B inataurault P Hier. in 01862,24 Caariotlor. 9 (Chr. m. II 120) Aug. tl.c.tl. XVII12 (210, 11) Draa. II 2, 1 (83, 2)

III 14

Fig.2

F (215F)

A

A Alexandrinorum

VI

G

VII

a

C - CXXXV· Olymp ·

D

Consules

B

Quintus Ennlus poet& Tarenti naseitur. Qui a Catone quaestoVIII re Romam translstus habitaVIlli uit in monte Auentino parco ad.... MDCCLXXX X modum sumptu contentus et uniOlymp . UB aneillulae miniaterio• ... H XI G _,. • Uirgo Ueetalis Romae a seruo COrrupt& propria 88 .manu in• 10 XII terimit XIII

c - cxxxvr:

B 16

D

XII II C .... CXXXVII · Olymp · XV G .... • ludaeorum pontifex Simon,

Oniae filius, clarus habetur. Sub quo Ieeus. filius Sirach, XVI Sapientiae librum componens, XVII liO XVIII quem uocant Panir.reton, etC .... CXXXVIII . Olymp. lam Simonis in eo fecit menXVIIII tionem _,.MDCCXC XX G - " XL· ferme milia Gallorum a Romania caeea 16

B

XXI

I

A

133

Syriae et Asiae

Macedonum

VI

III

VII

!III

VIII VIlli X

VI VII

E

...,

Uta. Obr.

v

ne 118 117

VIII XI H6 VIlli XII 135 IH XIII X Macedonium Antigonus . ann· XV. I ns Xliii

XV

II

XVI XVII XVIII

III !III

Ill

131 180

v

XVI III VI XX VII Syriae Seleucus Ceraunos . ann . III· I VIII

1119 1119 llt7

0 ad CXXXVI 01. L, ad XIII B a ad V 0, aed corr., ad VI A, ad CXXXV 01. LM • infra XIII A, ad CXXXVII 01. LPN tJ ad XVIII B, ad CXXXVIII 01. L, ad XVIIII OM

J

Syr. 24 ayrie L syria M airiae P eleuchus B 26 ceraunus BM annia B an· AP{OLM Pa.-lntlor. oUr. Allf1. e.195 (Chr. m.II 499) Bulp. &tl. ehron. I/19, 3 Hac. 13 tigonus Bantigonus (o. Rd. Atthi) N an· AP.B annia N {L

K

a) (Arm•..- N 3ellllius; ennius M 4questoreLPquaeetore• (m) A 5 habitabitOL 7 anciliae OBM Plr' ancilie L; aneillula fiiJl. Hier. ep. 6,11: 71,3: 107,9,2; 108,15 s.U.oa Suet. p.24 RsiUeraeh. b) Arm. 9 uestalee B 10 interemit M interimit (o. Rd. Atl•emit) N c) Arm. 16 iudeorG. N symon L aim B 18 hieeus ANLM ieeu o. Hd. I'M'b. 0 11irac0 syrach LB 19 conponens LBM paraneton 11erb. L 21 symonem IIM'b. -is L aymonis P in eo (0 .EeclmGBI. 50,1 Hier. comm. in Don. 9 ( P. L. XX Y 545d) Prorp, 251 (Chr. m. I 400) c:hron. Gall. o. 511 (Chr. m. I 637, 299) Intlor. iufl. 905 (Chr. m. Il450) BJ. 991 (Chr. m. III 911) lord. Rom. 75 d) Arm• ..- 23 quadraginta M 24 ceeum L caeae Jf ceea P Prorp. 954 (Chr. m. 1400) Oros. I Y 13,10 '\

L

M

III AN INTRODUCTION TO JEROME'S CHRONICLE AND A GUIDE TO ITS USE

15

Key to Figure 2 A- Fila regnorum: the vertically-listed regnal years of the kings whose people or kingdoms are listed at the top of the column. The synchronism for a single calendar year is achieved by reading across the fila horizontally. B- spatium historicum: the space where historical events and individuals are described. The regnal year opposite the first line of each entry is usually the year to which that entry should be dated: e.g. Ennius' birth belongs to 240, not 240-37. C- Olympiad markers. These note the year of each Olympiad, which is the year immediately following the marker. D- Years since the birth of Abraham. These are noted every decade and are underlined, as are all figures in that year across the page. E- modern years (in this case, BC, 'ante Christum'). These are the modern equivalents of the regnal years, and were added by the editor, R. Helm. F- concordance to parallel page in the edition ofJ.K. Fotheringham (London 1923). G- alphabetic markers, by which each entry should be cited, e.g. 133• (since this is page 133) for the entry on Ennius. H- an asterisk to note that this entry was added by Jerome and did not appear in Eusebius' original. I- rubrics announcing new kings. Each notice gives the name of the people or kingdom, the name of the king, and the length of his reign. Usually the king's position in the sequence of kings is given as well, though not here. In some instances additional material is included. Such rubrics are almost always followed by regnal year 'I'. J- apparatus criticus noting relative position of the historical entries with respect to the regnal years (in Latin). K- apparatus criticus for the rubrics (I, above) (notes in German). L- regular apparatus criticus (notes in German). M- indicates that the entry appears in the Armenian translation. N- indicates that the entry does not appear in the Armenian translation.

Sigla not represented in Figure 2 (*)- after an entry indicates that Jerome has modified an entry written by Eusebius, usually by adding material to it ( c£. H, above).

t- after the entry letter in apparatus 'L', above, indicates that the entry appears in the Syriac

epitome of Eusebius (see the apparatus to 20" and gin Figure 1 for examples). ' -)', '-) zu', 'Zu', ' - Zu', or'- zu' followed by a date 'Abr.', a German notice, and a'+ Arm.'in apparatus 'L' indicates that an entry contained in the Armenian translation was missed by Jerome. Figures 1 and 2 are reproduced with the permission of Akademie Verlag GmbH, Berlin.

III 16

He chose Abraham, chiefly because he was regarded by Eusebius and other Christians as either the first Christian or as a proto-Christian. The Chronicle is, therefore, a history of the known world since the first coming of Christianity, and his 'ann. Abr.' chronology is therefore a proto- AD system. Most scholars who mention the Chronicle cite the text strictly from these 'Years of Abraham' since in the past there was no other obvious way to cite individual years. Indeed, Schoene printed every Year of Abraham to assist citation, and until Helm's edition of 1913 all chronological figures in the margins and fila were printed in Arabic not Latin numbers, a practice that also simplified citation in this manner. Unfortunately Helm's fidelity to the Latin original (i.e. noting Years of Abraham every decade in Roman numerals) makes this method of citation very difficult. Some different methods of citation have been noted above, pp. 5-6. The few modern literary scholars who cite Jerome from Helm's edition usually just cite the page number followed by an 'H' (e.g. p.l71 H), if they bother to cite the entries at all. Ramsey and McGushin, noted on p. 5, above, make no attempt to cite the entries at all apart from the Years of Abraham and Olympiads, and McGushin unfortunately does not even indicate which edition he used. As can be seen from p. 20, Helm has placed a letter of the alphabet at the beginning of each entry. These entries occupy what is known as the spatium historicum, which runs down the middle of each page between the fila regnorum. On p. 20 those entries associated with the filum of the Assyrians are labelled a and c, those with the filum of the Hebrews are b and d, those with the Sicyonians are e and g, and that with the Egyptians is r. Usually there is only one main column of text in each spatium historicum (see Figure 2), but down to p. 105, and perhaps in other detailed sections of the Chronicle, Eusebius used multiple columns within the spatium, which Jerome often seems to reproduce (e.g. pp. 29b, 3lb, 40•, 43b, 46b, 47, 49b, 53 b, and 64 b). 40 Any entry should therefore be cited by its page number and superscript letter thus: 20bf (or just 20f since there is only one alphabetical sequence per spread). Each line is numbered on the inside of the double-page spread (there are twenty six lines per page, which reflects Jerome's original) for specific reference to lines (e.g. as in the apparatus) or to items that are not labelled with letters, chiefly the introductory material for each new reign (originally in the form of rubrics). Thus the material describing Ninus would be cited as 'p. 20a.l-6'. There is no need to use any other method of citation. At the bottom of the page we have a tripartite apparatus criticus. Contrary to normal practice the language of the apparatus is German (in italics), not Latin (except, for some reason, in the first apparatus). The first shows the relationship of the individual entries in the spatium historicum to specific regnal years. On p. 20a we can see that MS 0 is lacking for these early pages of the chronicle (to 47a, in fact) and that in MS B entry d starts opposite 'VI' and in MSS ALit starts opposite 'X'. This means that in MSS MPN (which are not cited) it starts at 'VII', where Helm has placed it. The text of the accession rubrics appears in the second apparatus and is referred to by the abbreviated name of the kingdom in bold, in this case Ass., Hebr., Sic., and Aeg., and by line number. In some cases, for additional support odor parallel 40 It is usually claimed that all Eusebius' entries were written like this (see Helm's edition, pp. xxvii-xxxiii, and Mosshammer (n.l) 62-3, 81-3), but I see no compelling evidence to support this claim.

III AN INTRODUCTION TO JEROME'S CHRONICLE AND A GUIDE TO ITS USE

17

statements, Helm also cites other witnesses to Jerome's text. These are set in italics. Here he cites eleven later witnesses to Jerome's rubric concerning Ninus and his regnal years. Finally we have the apparatus proper, which goes in alphabetical order of the entries and by line number. Again, later witnesses to Jerome's text are cited in italics. A bold dagger (t) after an entry letter indicates that that entry appears in the Syriac epitome of Eusebius' original Greek chronicle that appears in the Chronicleof724 (seen. 33, above). Later on in the text thissiglum is joined by one of two others, Arm. and , which indicate that a particular entry either is or is not present in the Armenian translation of Eusebius' Greek original (seen. 32, above). These are important for establishing the text and chronology of Eusebius' original work (see pp. 24-5, below), though, strangely, they are never actually quoted anywhere. One also occasionally finds'-)','-) zu', 'Zu', ' - Zu', or'- zu' followed by a date 'Abr.', a German notice, and a'+ Arm.' (e.g. 43a between d) and e); 78b between c) and d); and87batend). This indicates an original entry that has been missed by Jerome and appears in the Armenian translation. It is quoted from the German translation of Karst (seen. 32). One can see how this format of fila and spatium continues over the following pages. Because the Hebrew filum is on the left hand page and the Greek filum is on the right (Eusebius did not know much about the Assyrians or the Egyptians), one finds that Biblical history appears in the spatium historieum on the left-hand page and material from secular history appears in the spatium his tori cum on the right-hand page. Jerome did not adhere rigorously to this format and filled in many events from Roman history on the Biblical side when he did not have room on the secular side, and even Eusebius put secular material on the left when he did not have room on the right or if it related directly to a kingdom whose filum appeared on the left. As one skims through this first section of the chronicle an incongruous aspect that one immediately notices is that Euhemerized figures from Greek mythology cavort opposite figures from the Bible as part of the same 'reality'. In an apologetic twist, Eusebius makes it seem as though Biblical history is fixed and exact while pagan history is full of uncertainty and contradiction. He does this by offering multiple chronologies for the same people and events, such as Hercules, Homer, Hesiod, and the foundation of Carthage. 41 Page 105, the second year of Darius and the beginning of the reconstruction of the Temple in Jerusalem, marks the last of the double-page spreads. From page 106 Jerome allots one spatium per page and consequently the modem dates appear on the outside edges of both the left and right pages as the text shifts to single pages for the rest of the chronicle. Over a space of eighty-six double-page spreads and a further eighty-two single pages the fila ebb and flow, there being as many as nine per page (pp. 83-6) and as few as two (pp.l0715, 122-5, Persians and Macedonians; pp. 150-5, Alexandrians and Judaeans; pp. 164-87, Romans and Judaeans ), though usually there are six, seven, or eight at a time. The polyarchy of the past gives way to the monarchy of the Roman Empire, and with the capture of Jerusalem by Titus in AD 70 (p. 187), but a single filum remains, that of the Roman emperors, 41 See esp. 66a" and 86ad, and e.g. 26b1, 40be, 50bd, 53bd, 55bh, 56b1; Hercules: 40bh, 4 3bk, 49b 1, Slbh, 56b', 56b1, 57b\ 59b\ 59bk (= 40bh), 60bd (=1574-1196 BC); Homer: 63bd, 66a", 69b1, 7lbb (= 1160-l017BC) Qerome adds his own dating at 77b'); Hesiod: 7lbb, 84b', 87b1 ( = l0l7-767BC); and Carthage: 58be, 69be, 7lb', 8lb (= 1214-850 BC). The same argument against Greek history is made by Josephus, Against Apion, 1.15-27, 37-8.

III 18

and it continues down to the end of the work. From the capture of Jerusalem we have only the regnal years of the Roman emperors every year, Olympiads every four years, and Years of Abraham (underlined) every ten years down the left hand side of the page. At '20 Constantine' Jerome marks the point at which Eusebius concluded his original chronicle and he takes up the continuation (231£). The fila and spatia conclude on p. 249 with year fourteen ofValens and Valentinian (378) (the '[XV]' should be ignored) and the text itself concludes on p. 250 with a final chronological supputatio accounting the years covered by the chronicle from its conclusion backwards to the seven chronological linchpins (these are discussed below, pp. 20-1). Following the supputatio on p. 250 are three pages of apparatus aitici that were too long to fit at the bottom of preceding pages (as mentioned above, Helm maintainedJerome's original twenty-six-line page format), the index of names (pp. 254-78), and the testimonia of the witnesses and parallels (pp. 279-455). In this latter section close or exact parallels are quoted and close parallels from Eusebius or Jerome's sources (or witnesses to their sources) are cited with spaced text (e.g. Josephus for Eusebius and Eutropius for Jerome). 4.2. The Fila in Detail I. Assyrians. Begins in 2016 BC with '43 Ninus' (20a) replaced by the Medes in '20 Sardanapallus' (820 Be) on p. 83a.

2. Hebrews. Begins in 2016 BC with the birth of Abraham. On p. 72a in 996 BC it splits into the Hebrews called Juda, on the left side of the left spatium historicum, and those in Samaria who were called Israel, on the right side. The filum for Israel ends with the capture of Israel by Sennacherib, the king of the Chaldaeans, in 747 BC (p. 88a). The filum for Juda becomes an accounting of the seventy-year Babylonian captivity ('Iudaeorum captiuitas') starting on pp. 99a-100a (590 Be). This filum ends in 521 BC on p. 105a. A new filum starts in 160 BC for Iudas, the 'ludaeorum dux', a position that is elevated to 'rex' in 130 BC (pp.141 and 145-6). With the Roman conquest of Palestine by Pompey the filum for the Judaean kings shifts to the High Priest Hyrcanus in 67 BC (p.153) and then to Herod in 33 BC (pp. 160-2). The filum finally ends in AD 70 with the capture ofJerusalem by Titus (p. 187), after a further succession of a dux, a tetrarch, a princeps, and a rex. 3. Sicyonians. Begins in 2016 BC (p. 20b) and ends in 1129 BC (p. 65a). 4. Egyptians. Begins in 2016 BC (p. 20b) and ends on p. 104b in 525 BC with the capture of Egypt by Cambyses, king of the Persians. It reappears in 413 BC (p.116), but ends again in 350 BC (p.121). 5. Argives. Begins on p. 27b with Inachus (1856 BC), and ends on p. 53b with King Acrisius (1313BC). 6. Athenians. Begins with Cecrops, the first king of Athens, in 1556 BC (p. 41b) and comes to an end in 684 BC with the end of Eryxias and the monarchy (p. 93a). 7. Mycenaeans. Begins in 1306 BC42 with Eurystheus (p. 54b) and ends just after the capture of Troy under Agamemnon in 1179 BC (60b ). 42 The first two Roman numerals have been shifted down two lines because of the length of the rubric and so it at first looks like 1304 BC.

III AN INTRODUCTION TO JEROME'S CHRONICLE AND A GUIDE TO ITS USE

19

8. Latins. Begins with Aeneas on p. 62b (1178 Be) and on p. 88b Romulus appears as the sixteenth king of the Latins as well as the first king of the Romans (752 BC). 9. Lacedaemonians. Begins in 1101 BC with Eurystheus (p. 66b) and ends in m BC with Thalcamenes (p. 86b). 10. Corinthians. Also begins in 1101 BC, with Aletes (p. 66b ), and ends in 779 BC with Automenes (p. 85b). 11. Medes. Replaces Assyrians in 819 BC on p. 83a under Arbaces and is replaced by the Persians in 561 BC on p. 102a, when Astyages is king. 12. Macedonians. Begins on p. 83b with Caranus (813 BC) and continues down top. 140, where it ends in 167 BC under Perses (Perseus). 13. Lydians. Begins with Ardysus in 778 BC (p. 85b) right after the end of the Corinthians and ends in 548 BC with the capture of Croesus by Cyrus (p. l03b ). 14. Romans (kings). On p. 88b Romulus appears as the sixteenth king of the Latins as well as the first king of the Romans (752 BC). With the expulsion ofTarquinius Superbus in 513 BC (p. 106) the filum for the Roman kings comes to an end and the spatium historicum, which previously had not been labelled at all, henceforth has the headings 'Romanorum' on the top of the left-hand pages and 'Consules' on the right-hand pages. There is no evidence that Eusebius (or Jerome) ever included any of the names of the consuls, though Jerome must have used a consular hst as a source to correlate his historical material (see below). 15. Persians. Replaces the Medes on p. 102a (560 BC) and ends in 330 BC (p. 124) with the defeat of Darius by Alexander the Great. 16. Alexandrians. The Ptolemaic kingdom of Egypt begins in 324 BC (p. 125) with Ptolemy I and ends in 29 BC (p. 163 ), even though the deaths of Antony and Cleopatra are recorded in 33 BC (p. 162), the result of the 'corrections' by Jerome noted above, on pp. 4-5. 17. Asia. Begins withAntigonus in318 BC (p.126) and is merged with Syria in 283BC (p.129). The joined filum ends in 93 BC (p.149). 18. Syria. Begins with Seleucus, king of Syria and Babylon, in 312 BC (p. 126), which then merges into a single filum (Syria and Asia) in 283 BC (p. 129). The joined filum ends in 93 BC (p. 149). 19. Romans (emperors). Julius Caesar is counted as the first Roman emperor and his regnal years begin in 48 BC just before the Battle of Pharsalus. The spatium historicum is no longer headed 'Romanorum consules'. Normally each rubric for each new king is written in a narrow column just above its own filum. The rubrics for the Roman emperors, however, stretch right across the page, filling the entire space once occupied by the various diverse kingdoms of the past (the Years of Abraham and Olympiads still follow along in the left margin). The filum for the Roman emperors now takes the primary position running down the left-hand side of the page. The fila for the Alexandrians and the Judaeans (the only fila left at this point) are squeezed over to the right-hand edge of the page as the spatium historicum increases in size, and the regnal years are listed further apart to take in the extra historical accounts: for example, pp. 134-5 cover twenty-six years (225-200 Be), 154-5 cover sixteen years ( 64-49 Be), and pp. 162-3 only six years (33-28 Be).

III 20 4.3. The Chronological Linchpins Eusebius began his researches by establishing seven chronological points on which he could hang his overall chronology. These were famous and well-established dates that he could use to work other more poorly established chronologies backwards and forwards. It also gave him definite blocks of time in which to fit other events of the same fila, for instance the number of years between the capture of Troy and the first Olympiad. A number of these linchpins were valuable for coordinating the chronologies of different fila, especially those of the Hebrews and the various Greek kingdoms, or fila like the filum of the Assyrians that could be coordinated with the Greek fila. The first linchpin, of course, as we have already seen, was the birth of Abraham in 4 3 Ninus. The accession of King Cecrops in the thirty-fifth year of Moses marks the second linchpin: the synchronism of Moses and Cecrops. In 41ac Eusebius notes that there are 375 years from this year to the capture of Troy. At 43ad he notes the exodus from Egypt in 45 Cecrops and states that from this year to Solomon and the building of the Temple there are 480 years. Both of these dates will be referred to later in the chronicle. On pp. 60-1 we meet the third chronological linchpin, the capture ofTroy in 835 Abr. ( = 1182 BC), 410 years after the birth of Moses, 375 years after the first year of King Cecrops of Athens and the thirty-fifth year of Moses, and 406 years before the first Olympiad. On p. 70 (esp. 70a) we meet the fourth linchpin, the inception of the construction of the Temple in Jerusalem by Solomon, 480 years from the exodus from Egypt. From the Flood to Moses Eusebius calculates 1,447 years, and from Adam to the Flood 2,24 2 years, which to the present year (1033 BC) makes a total of 4,169 years. The first Olympiad, the fifth linchpin, appears in 776 BC (p. 86). From this point Jerome marks every fourth year with an Olympiad in the form 'I·Olymp.', 'II·Olymp.', and so on, the actual year or anniversary of the games falling in the year immediately following the Olympiad marker (in this case, 776 BC or 1241 Abr.). This is important and will be discussed later (pp. 22-4 ). The sixth chronological linchpin, the inception of the rebuilding of the Temple in the second year of King Darius (521 BC), appears on p. 105, the last of the double-page spreads. This is a key date for it precisely synchronizes Greek, Persian, and Biblical history. In the fifteenth year of Tiberius, AD 28, we meet the seventh and last of Eusebius' linchpins, the beginning of Christ's ministry, 548 years from the rebuilding of the Temple, 1,060 years from Solomon and the first building of the Temple, 1,539 years from Moses and the exodus from Egypt, 2,044 years from the birth of Abraham. From the flood to Abraham there are 942 years and from Adam to the Flood, 2,242 years (pp. 173-4). It is interesting that Eusebius chose Christ's ministry rather than crucifixion as the linchpin here. It was probably because there was no agreement as to the date of the crucifixion, whereas the beginning of his ministry was explicitly dated by Luke 3:1 (or so it was thought; it actually dates the beginning of John's ministry).

III AN INTRODUCTION TO JEROME'S CHRONICLE AND A GUIDE TO ITS USE

21

The final supputatio (p. 250 ), mentioned above on p.l8, gathers all these linchpins together and calculates the number of years elapsed from the end of the chronicle (AD 378, the end of Jerome's continuation43) to those points: from Tiberius' fifteenth year and the ministry of our Lord Jesus Christ: 351 years from the second year of Darius, king of the Persians, at which time the Temple was rebuilt: 899 years from the first Olympiad, at which time Isaiah was prophet amongst the Hebrews: 1,155 years from Solomon and the first building of the Temple: l,4ll years from the capture of Troy, at which time Samson lived amongst the Hebrews: 1,561 years from Moses and Cecrops, the first king of Attica: 1,890 years from Abraham and the reign of Ninus and Semiramis: 2,395 years Furthermore, from the Flood to Abraham: 94 2 years and from Adam to the Flood: 2,24 2 years From Adam to the fourteenth year ofValens, that is to the consulship ofValens for the sixth time and Valentinian for the second, this makes a total of 5,579 years There are a number of other scattered chronological markers used in the Chronicle: and it would be best to mention them here, since they are not explained anywhere and will puzzle unfamiliar readers. The first is the 'Iobelaei secundum Hebraeos' (jubilees according to the Hebrews'), which is an accounting ofJewish universal chronology from the creation of the world. AJubilee is a unit of fifty years and Eusebius andJerome mark the beginning of every tenth Jubilee from the Jewish date of the creation of the world, the implication being, of course, that the Jews were wrong. The first notice is the beginning ofJubilee 41 at 22a", thus marking two thousand years from Adam, though by Eusebius' reckoning it is 3,228 years. The notices for Jubilees 51, 61, 71, 81, and 86 appear at 46ah, 73ah, l09n, 174", and223h. These are not to be confused with the 'Iobelaei a maioribus' or 'iuxta maiores nostros' or 'secundum maiores'. It is not known what these are or who the 'maiores' are. They have no sequential numbers as the Hebrew Jubilees do and they start in 204: 'in hoc anno iobelaeum a maioribus inuenimus obseruatum, id est XII anno Seueri et CCLI 44 Antiochenae urbis' (212h). This date strongly that they take their origin from the Secular Games celebrated by Severus in 204. 4 They are noted every fifty years after this, 219e, 22i, and 238i, though Jerome mistakenly labelled the last as 'secundum Hebraeos' instead of 'secundum maiores'.

43 Eusebius' supputatio in 325 was omitted by Jerome. It survives in the epitome in the Chron. 724 (see above, n. 33). 44 This should be Year 252. 45 See A. E. Birley, The African Emperor: Septimius Severus (London 21988) 156-60.

III 22 5. Determining the Chronology of an Entry 5.1. The Simple Method For the most part there is a very simple way to establish the date Eusebius or Jerome assigned to any entry: it is the BC ('a. Chr.') or AD ('p. Chr.') date Helm has provided opposite the regnal year under which the first line of the entry falls. If the entry extends through one or more regnal years (described below), it is usually the regnal year of the first line of the entry, where the superscript letter is, that marks the correct year, though there do appear to be a few exceptions where Jerome was prevented from starting long entries exactly where he wanted because he could not shift or compress Eusebius' entries sufficiently (e.g.l33a dates to 238 BC and Btl dates to 185). When an entry starts at an Olympiad it is normally to be taken with the following regnal year. 5.2. The Complicated Explanation Of course, the reality is rather more complicated Eusebius' year, that of Caesarea, like most of those in the eastern Roman empire, was based upon the Macedonian calendar. Analysis of Eusebius' correlations between his own calendar and the Roman calendar in his Martyrs of Palestine shows that his calendar was of the type used inTyre and therefore probably began on 3 October. 46 Olympiads ran from July/August of each quadriennium, though for chronological purposes historians and chronographers divided each Olympiad into four calendar years. Each regnal year normally began on the anniversary of the day each king or emperor ascended the throne, except for those such as the Egyptian and Ptolemaic regnal years ,for which the beginning of the second regnal year was always calculated from the next New Year's (however short the first year), so that regnal years and calendar years were almost always in synchronization. The calendars of some kingdoms were solar, some were lunar, and some were lunisolar; consequently they were never in synchronization with one another from year to year. 47 The Years of Abraham were merely notional units designed to give a relative chronological context for the work as a whole and simply counted each passing year from 2016 BC. All these differing calendar systems would have been impossible to calculate, synchronize, and record if they were all treated as accurate representations of reality. In order to work they all had to be coordinated to a single system. Thus all years of Abraham, all regnal years, and all Olympiads were equated with the civic Macedonian calendar year of Caesarea. The various regnal years that are relevant to the suicide of the Vestal virgin in Figure 2 (133b), for instance, 'XI Alexandrinorum', 'XI Syriae et Asiae', and 'VIII Macedonum', as well as the year 1781 Abr. and Olympiad 136.1, were all equated exactly with 3 October 237BC to 2 October 236 BC by Eusebius, regardless of the exact period of time each represented. For all his history from 776 BC, Olympiads were Eusebius' primary chronology, and the correlation of the Olympiads and Macedonian months already existed in his chief source, an 46 J.-P. Rey-Coquais, 'Le calendrier employe par Eusebe de Cesaree dans les Martyrs de Palestine', AB 96 (1978) 55-64, esp. 62 (1 Hyperberetaios • 3 October). For the detailed argument, see Burgess (n. 26) 28,104-6. 47 On ancient calendars, see EJ. Bickerman, ChronologyoftheAncientWorld (London 21980) and Samuel (n.15). See also Burgess (n. 26) 28.

III AN INTRODUCfiON TO JEROME'S CHRONICLE AND A GUIDE TO ITS USE

23

Olympiad chronicle. 48 That the ria Augusta. Of these, Ammianus' history, the Epitome, and the HA had not even been written when Jerome was working on the canones and these pagan authors are hardly likely to have even known about his Christian chronicle, let alone used it, especially since it provided no pertinent or useful information that was not easily available in other secular sources, that is consularia/fasti and, as we shall see, the KG. The key point, however, is that in his preface Jerome describes the canones as a tumultuarium opus and says, notario ... velocissime dictaverim. Such conditions were hardly conducive to the use of multiple sources for single entries. He obviously had no time to peruse four or five works at a time and combine bits and pieces from a number of them for each of his brief notices. The conclusion must be that for these entries Jerome was using the KG alone, not a combination of Eutropius and seven other works, three of which had not even been written, since the KG is acknowledged to derive ultimately from Suetoni us in the early lives (see above, p. 112) and to have been a source used by six of the remaining seven. Although the author of the KG did not use the Chronica urbis Romae, the two authors appear to have drawn on the same sources for their information on buildings. For the period from the late third century to 337 Jerome's sources for nonreligious information were, then, Eusebius' Chronici canones to 325, the KG, an early recension of the Consularia to 370, and a now lost Antiochene continuation of Eusebius that extended from 326 to 350 (the Continuatio Antiochiensis Eusebii). 42 Usually the material copied from these sources is separate and discrete, but on the odd occasion when two sources gave different information for the same or a related event Jerome did combine them, but almost always in separate blocks. 43 In only two cases is there complicated and intricate interlace: 226e ( = A.D. 298), concerning Diocletian's actions taken against Egypt, where material from the KG was interpolated into an entry from Eusebius, and 228d (= A.D. 305), where, influenced by the Consularia, Jerome changed imperium to purpuram. 44 After the conclusion 41. Part of the famous Chronograph of 354, ed. Th. Mommsen, Chron. min. l: 145-48, a pagan work concluding with the death of Licinius in 325, yet pointedly omitting Constantine. On this work, see now Michele Renee Salzman, On Roman Time: The Codex-Calendar of 354 and the Rhythms of Urban Life in Late Antiquity (Berkeley, 1990), 52-56. 42. What follows derives from my "Jerome's Chronici canones: Sources, Chronology, and ComposiseeR. W. Burgess, "The Chronici canones of tion. A.D. 282-378," which is in preparation. For Eusebim of Caesarea: Content and Chronology, A.D. 282-325" (forthcoming); for the Consularia, see idem. The Chronicle of Hydatius and the Consularia Constantinopolitana: Two Contemporary Accounts of the Final Yean of the Roman Empire (Oxford. 1993). 175-209, esp. 196-97; for the Continuatio Antiochiensis, see idem, The Continuatio Antiochiensis Eusebii. A Mid-Fourth Century Antiochene Continuation ofEusehiw' Chronici canones (forthcoming). which includes a text and translation of the hypothetical reconstruction. 43. Entry 225a (=A.D. 287), of which the first sentence is from Eusebius and the second from the KG; 229' (= 308), where Carnunti has been added from the Consularia to the KG; 229h (= 311) and 229' (= 312), where apud Tarsum and iuxta pontem Mi/uium have been added to Eusebius from the KG; and 231' (= 326) of which Vicennalia ... acta is from Eusebius and et . .. edita is from the Consularia. 44. 226': (Alexandria cum) omni (Aegypto per Achilleum) ducem (a Romana potestate desciscens) [octavo obsidionis mense a Diocletiano] capta est. Itaque (plurimi) per [totam Aegyptum gravibus proscriptionibus exiliisque vexati (interfectis)] his, (qui auctores perduellionis extiterant); 228d: (Secunda anno

*

V 124

of Eusebius, between 326 and 337, there are no examples of similar combinations of sources. 45 Thus to 337 Jerome's method of compilation is quite clear and consistent, and all his historical material is completely accounted for. However, if one assumes that the KG concluded in 337 and that Jerome replaced it with Eutropius (since his similarity to Eutropius continues), in the period after 337, where he should only be using the Continuatio Antiochiensis, Eutropius, and the Consularia as nonreligious sources, one suddenly finds him providing details that cannot be attributed to any known or postulated source, often simply the geographical location of an event. 46 Some of these details appear in Aurelius Victor or the Epitome, two works that also used the KG. Furthermore, Jerome's apparent combining of Eutropius, the Consularia, and this unknown source is unusual and rather complex. 47 A number of Jerome's deviations from Eutropius find parallels in the Epitome, a work that would not have used Jerome as a source and that provides in close association with the parallels to Jerome additional information not contained in Jerome or other extant sources. 48 One is faced with a choice: either Jerome drastically changed his method of composition when the KG ran out in 337; he began copying Eutropius; he obtained another, unknown source that he used sparingly; and his chronicle was later used as a source by the pagan author of the Epitome, who possessed another unknown source that provided information very similar to but fuller than the material in Jerome, or Eutropius, Jerome, and the Epitome were still using the KG. The choice seems obvious. In most cases below Jerome seems to be copying Eutropius but includes material that does not derive from Eutropius, or his other two nonreligious

persecutionis) [Diocletianus Nicomediae, Maximianus Mediolanii] Sui/a et Othone Silano et Othone Silano et Antonino}, though for some reason, someone later changed Othone in the Descriptio to Catone (in both places) and further corruptions have made a complete mess of the years 51-53 in the FVprlpost!Prosper(§§ 137-41 and431-5). 63 The error is very early in the tradition since it also appears in the Chron. Pasch. (348. 7-8). 64 In the Descriptio, instead of Geminis we find Si/anis, which has been miscopied from Si/ano et Nerua in the previous year. Epiphanius retains the original llllo ft!ltVOOV (292.11) as does the Chron. Pasch., slightly modified to ft11ivou Kotl ft11ivou (p. 389.11). FVpr has just duobus Geminis, while FVpost has Ruffio Gemino et Rebel/io Gemino, like Prosper(§ 391, Fufio Gemino et Rubel/io Gemino}, through influence from other sources (because of the fame of this consular year as the date of the birth of Christ; see below n. 67). 65 The Chron. Pasch. presents each of these in an expanded form Kotl AtvtollA.ou (p. 366.3), Kotl (p. 388.3) and ft11ivou Kotl ft11ivou (p. 389.11), though the list ofEpiphanius, which was written c. 375, extends from 2 BC to AD 29, and derives directly from an earlier translation of the Descriptio than that used by the author of the Chron. Pasch. (see Hydatius [cit. n. 8], 197}, still preserves the t&v Iiilo Ul;trov and Iiilo ftlltv&v of the Descriptio (291.2 and 292.1 1). The Chron. Pasch. and Epiphanius still retain the geminations for each year as well (CP 366.5, 388.5, 391.20; Epiph. 291.3 and 292.11-13).

66 Even the list used by Malalas in the sixth century had duobus Sextis, which he rendered as Ul;tou Kotl (10.7, p. 232.11}, exactly like the expanded duobus forms in the Chron. Pasch. (see nn. 64 and 65). This parallel, that noted above in n. 52 (Rusticio for Etrusco), and the common errors in AD 17 ('Rufinus' for 'Rufus', at 236.16) and 68 ('Tolpillianus' for Trachalus at 258.1 [cf. FVpr, 'Turpilio'; Prosper, 'Turpilianus'], where he has compared this list to another that had 'Trochelus') demonstrate that Malalas' Jasti also descended from the same precursor as FVpr/post!Prosper. On the other hand, Malalas repeats the bizarre error of the Descriptio in 2 BC of naming Octavian rather than Caesar or Augustus (227.7-8}, and like the Descriptio, calls Vinicius 'Vindicius' in AD 2 (230.14). Perhaps the comparison Jasti was related to the precursor of the Descriptio (which has 'Trabalus' in 68).

XV 276 However, I have not found any instance of this form in a written source earlier than 161 -though Tacitus does describe this year as Rubellio et Fufio consulibus, quorum utrique Geminus cognomentum erat (Ann. 5.1.1)- and it may be that all these unusual references to the year of Christ's birth ultimately derive from the Fasti ofl6J.67 As we have seen above in section 111.2.1, the form duobus Sextis in AD 14 also appears in the Chron. 354, which has Lentulo et Lentulo and Gemino (!t Gemino in the other two instances. The key to understanding the gemination, however, is Suetonius, who gives the consuls of AD 14 as duobus Sextis, Pompeio et Apuleio cons (Aug. 100. 1), the peculiar treatment of the praenomen showing a clear link among Suetonius, the Chron. 354, and theFasti of 161 (see Table 1). IV.4. The Duobus Formula The text ofthefasti used by the compiler of the Fasti of 161 therefore already had one entry, that of AD 14, in the form duobus Sextis, Pompeio et Appuleio. This was part of a text that had at one point extended to some date between 78 and 92. I have found only one epigraphic and two painted consular dates that refer to contemporary or near contemporary consuls with this duobus formula.68 All three are private records; none is what we could call 'official' or 'public'. In a similar vein, Augustus referred to the year 18 BC in his Res Gestae as Cn. et P. Lentuli (6, Greek only/reverse order, and 18, Greek and Latin) and a formal inscription containing consular dates between 140 and 172 has Quintiliis cos (151) for Sex. Quintilius Condianus and Sex. Quintilius Valerius Maximus (CIL 14.246 11.8), the former example using cognomina, the latter using nomina. Finally, an exact parallel to duobus Sextis, Pompeio et Appuleio can be found on tombstones, where this formula was used when naming family members.69 It is clear, therefore, that there existed an unofficial tendency to condense similar names in this fashion in the Roman world. Finally and more pertinently, there is also evidence that M. Aurelius and L. Verus were in fact referred to during their reign as 'the two augusti': Imp duor(um) Aug Antonini et Veri Armeniacorum (CIL 7.1211 = AE 1952.88) and Pro salute et reditu et uictoria Imp duorum Aug (CIL 3.3432; a dedication of 164). Even more important, there are at least three examples of epigraphic consular dates from 161 with similar forms, Augustis n(ostris) cos (CIL 15.353 = 11.6684; a brick stamp), Imp Augustis cos (CIL 3.1295) and Augustis cos (CIL 12.5905), and one with exactly the form under consideration here, dating from 1 April, II Aug cos (CIL 6.126).70 Clearly this identification began early 67 Simply listing those in Mommsen's Chron. min. volumes, we have the preface to the Liberian catalogue of Roman bishops contained in the Chron. 354 (I: 73.2), the Computatio a. 452, 69 (I: 153.13); Prosper§ 388 (I: 409--10); Victorius of Aquitaine, Cursus Paschalis (I: 683.22, 686), and the Prologus Paschae ad Vita/em (I: 737.32). See also Lactantius, de mortibus persecutorum 2.1 and Div. Inst. 4.10.18; Augustine, de ciuitate dei 18.54 (ed. Dombart-Kalb, p. 344.3); and the Anonymi Libel/us de computo Paschali, PL 59: 553A and D (of the mid-fifth century), as well a third century reference from Ulpian in Mosaicarum et Romanorum legum collatio 8.7.3, in Paul Krueger, Theodor Mommsen, and Wilhelm Studemund (eds.), Collectio librorum iuris anteiustiniani 3, Berlin 1890, 166.15. Certainly the Liberian catalogue (seen. 30, above), Prosper, Victorius, and Augustine (seen. 80, below) all ultimately derive from the Fasti of 161. 68 There is a painted label on an amphora from 18 BC that records the consuls of that year as duobus Lentulis cos (CIL 15.4539). The tombstone of a soldier notes that he was discharged duobus Geminis (CIL 6.2489) and an amphora also has Geminis II cos (CJL 15.4573). 69 E.g. duobus C/audiis Moscho et Heliopo (CIL 6.14105), duobus C/. Cl. Faustino et Hermogenifilis [sic] dulcissimis (CIL 6.15053, where the C/. Cl. is an abbreviation for C/audiis), Tettio Hermeti seniori et Tettio Hermeti iuniori duobus Collibertis (CIL 6.27303), and duobus Terentis Baccho et Terentiano (CIL 6.27167), though the common name here (except for Collibertus) is the nomen not the praenomen or cognomen as in the consular dates. 70 The third example is an inscription celebrating M. Aurelius' birthday (26 April) and thus clearly belongs to 161. The editors incorrectly think that the II of the last example is an iteration and Mommsen dates the inscription to AD 70. The others must date to 161 rather than 202 (see below), since by 202 each emperor was represented with a separate final Jetter in an abbreviation, so that IMPPP A VGGG, for example, refers to three emperors. After the damnatio memoriae of Geta in 211 even the extra letters that represented him were removed from inscriptions (for an example, seen. 75, below). However, this system only began to develop later in the reign of Marcus and Verus, first with IMPP and only later with AVGG (CIL 7.504, optimorum maximorumque Impp n(ostrorum); 8.9698, Impp Antonini et Veri Aug; and 8.11173, lmpp Antonino et Vero

XV 'Non duo Antonini sed duo Augusti'

277

in the reign since the fact that they were two equal augnsti was the characterizing feature of their reign. IV.5. The Gemination of 161 The loss of Aug after each consul's name would have been normal for an entry such as this in the writtenfasti,71 but because of the unique example of two augusti in this case it was retained. Under the influence of the pre-existing duo bus Sex tis, Pompeio. et Appuleio, the tendency to refer to the new emperors as the duo augusti, and perhaps unofficial instances of the use of II Augus tis cos, the officially promulgated Imp Caesare M Aurelio Antonino Aug III et Imp Caesare L. Aurelio Vera Aug II became M Aurelio Antonino III et L. Aurelio Vera IL duobus augustis. Thus inspired, the compiler of the Fasti of 161 undertook similar reductions in 18 BC and AD 29, though in these cases the basis was the cognomina, Lentuli and Gemini. At a later date, when the names in a text descending from the Fasti of 161 were being reduced from three- and four-element names to single names, these forms were mistakenly treated as referring to separate consular pairs, duobus Sextis in one year and Pompeio et Appuleio in the next, though in the Chron. 354 the nomina in AD 14 were properly removed. This error arose because in manuscripts the names were usually crammed in together one after the other in double columns, not one line per year as in modern editions. The same happened in 18 BC, AD 29 and AD 161. At some date the original duobus Lentulis Cornelio et Cornelio in 18 BC was miscopied as duobus Lentulis Lentulo et Cornelio (Descr. 18b, FVpr 53), in the same way that duobus Geminis became duobus Silanis in the Descriptio in AD 29a, the Silanis having been copied from the consul of the previous year. These four geminations thus serve to identify the tradition of the Fasti of 161. V. The Later Duobus Forms There are two later examples of the duobus form in the writtenfasti, duobus Silanis in AD 189 and duobus Aspris in 212. These appear in all Latinfasti, even the Chron. 354, but with no geminations.7 2 The reason for this is that by 189 the duobus form had become an officially promulgated form. I have not yet been able to discover any inscription that records the consuls as duobus Lentu/is (18 BC) or duobus Sextis (AD 14), and only one each that records the consuls of 29 as duobus Geminis (CIL 6.2489 from Rome, noted above) and of 161 as duobus augustis (CIL 6.126 from Rome, noted above). However, of thirteen inscriptions for 189 that I have found, eleven give duobus Si/anis, and of fortyseven inscriptions and papyri for 212 thirty-seven give duobus Aspris.73 All thirty-nine laws dating from Augg). An amphora dated Ca]es n(ostris) cos may also belong to this year (CIL 15.4397). 71 There are no examples of the retention of Aug before the 360s and even after that it is very sporadic. 72 The Greekfasti of Theon give :!:tf.avo