"Adrien de Valois and the Chapter Headings in Ammianus Marcellinus" more

Notes and Discussions ADRIEN DE VALOIS AND THE CHAPTER HEADINGS IN AMMIANUS MARCELLINUS 233 In an important article in the hundredth volume of Classical Philology, Richard Burgess reexamines the relationship between various extant Latin sources for the later fourth century of our era. 1 He sets out significant parallels between Jerome, Eutropius, Festus, Ammianus, and the Epitome de Caesaribus for the period 358– 378, and attributes them to their common use of successive continuations of the lost Kaisergeschichte. The existence of this work, posited by Alexander Enmann in 1884 to explain the shared features of Aurelius Victor and Eutropius, is widely accepted, but it has not hitherto been argued that anybody wrote a continuation beyond 357.2 This is a bold conclusion, as Burgess admits, and significant in its implications, not least for his own earlier work on the Kaisergeschichte. 3 Further debate on several fronts seems likely. 4 This note has a narrower focus. In a speculative digression, Burgess observes that the chapter headings found in modern editions of Ammianus Marcellinus on occasion provide factual details and linguistic usages that are absent from the actual history, but that can be paralleled by Eutropius’ Breviarium, Jerome’s Chronicle, and (less markedly) the Epitome de Caesaribus; he suggests that their author, “agreed to be . . . a later editor and not Ammianus himself,” may have gained this additional information from the Kaisergeschichte. 5 I shall explain why this theory (which does not affect Burgess’ central argument) cannot be right, and add a few remarks on the characteristics, sources, and reception of the chapter headings. Modern editions of Ammianus do not enlighten their readers on the subject, and indeed sometimes mislead them, but the chapter headings are not to be found anywhere in the manuscripts. 6 The current divisions of chapters were devised, and the chapter titles composed, by Adrien de Valois (Hadrianus Valesius, 1607–92) for his edition of 1681. 7 Adrien presented his work as an editio posterior of the brilliant text and commentary that his late brother Henri de Valois (Henricus Valesius, 1603–76) 1. Burgess 2005. I thank Professor Burgess for encouraging me to publish this note. I also profited from the suggestions of CP’s referee and from discussion with Daniël den Hengst, Adrian Murdoch, Michael Reeve, and Roger Tomlin. Quotations from early modern books preserve the original capitalization, punctuation, and spelling; abbreviations are expanded. All translations are mine. 2. Enmann 1884. The terminal date of 357 is argued by Barnes (1970 and 1978), Bird (1973), and Burgess (1995). 3. Burgess (2005, esp. 189–90) markedly revises some arguments of his 1993 and 1995 articles. 4. Both Burgess’ list of parallels between the sources for the period (2005, 171–82) and his reassessment of the nature of the breviary tradition (187–90) are important innovations that demand a response. His vigorous dismissal of the lost Annales of Nicomachus Flavianus as a factor (168–69) will also be found provoking in some quarters. 5. Burgess 2005, 185–87, quotation from 185. 6. On the manuscript tradition, see Clark 1904; Robinson 1936; Reynolds 1983, 6–8; Broszinski and Teitler 1990. 7. A. de Valois 1681. All subsequent editors use the headings without explanation. Clark (1910–15) and Seyfarth (1968 and 1978) are silent. Rolfe (1935, 2 n. 1) attributes them to an early editor. Hamilton and Wallace-Hadrill say the same but wrongly imply that they appear in the mss (Hamilton 1985, 9). Only Galletier’s Budé (1968, 49) mentions without further comment, in a list of editions, that Adrien de Valois’ includes chapter headings. I cannot recall seeing the authorship of the headings discussed in any recent monograph or article, until briefly mentioned in Paschoud 2006, 242 (a response to Burgess 2005, published after this note was submitted and accepted). On ancient and early modern chapter headings in classical texts more generally, see Dionisotti 1997 and Schröder 1999. 234 Notes and Discussions had published in 1636. 8 Henri’s lengthy and learned endnotes were incorporated as footnotes, and were augmented by additional annotations that he had made in the forty years after 1636, as well as by occasional observations from Adrien. Notes from the 1609 edition of Friedrich Lindenbrog (Lindenbrogius) were included as an appendix, these too being bolstered by previously unpublished afterthoughts. Adrien’s fraternal piety was considerable (he had also composed a short life of his brother in 1677), and he modestly presented himself as carrying out his brother’s unachieved plans for a second edition. 9 For all that, his own contributions in historical, textual, and organizational terms were considerable. Adrien’s distinction as a historian—he was author, inter alia, of a three-volume history of the Franks—had been recognized by his appointment, alongside his brother, as a Historiographus Regius of Louis XIV, and he included in his Ammianus a lengthy endnote on the Hebdomon in Constantinople (so he identifies, both in this note and the chapter heading, what Ammianus in 26.4.3 calls simply a suburbanum). 10 It will be clear to any reader of Adrien’s introduction and notes that he improved the text significantly both by collation of manuscripts and by conjecture.11 And the introduction also explained various useful improvements that he had made to the presentation of the text: ac primum singulis Ammiani libris suam epitomen varia in capita distinctam praeposui, ut cuncti quid unoquoque libro contineretur, statim in prima ejus fronte notatum conspicerent: singula capita cujusque epitomes marginibus paginarum suo quodque ordine apposui: Imperatorum nomina, paria Consulum, et annos Christi nati singulis paginis, quantum fieri potuit, exacte praefixi, ut legentes statim scirent, sub quo quidque Principe et quo anno gestum esset. 12 And first, I prefaced each individual book of Ammianus with its own epitome divided up into various chapters, so that everybody could see immediately listed at its start what is contained in each book; the individual chapters of each epitome I placed in order in the margins of the pages [Caput I, II, III, etc.]. I placed at the top of the individual pages the names of emperors, pairs of consuls, and years of the Christian era with as much precision as was possible, so readers could tell immediately under which prince and in which year everything happened. These “epitomes” are the familiar chapter headings of modern texts, and though I shall continue to call them chapter headings, it should be observed that they were composed as tables of contents for individual books, which is how the editions of Clark and Seyfarth print them, rather than to be placed at the head of individual chapters, as in the Loeb and Budé editions. The theory that the author of the chapter headings used the Kaisergeschichte must be ruled out with his identification as Adrien de Valois. 13 If a scholar of his quality had gained possession of a lost text of that importance, he would certainly have recognized its value and published it (perhaps it would now be called the Anonymi 8. H. de Valois 1636. 9. A. de Valois 1677, 26–27, and 1681 (second and third pages, unnumbered, of Adrien’s praefatio). 10. A. de Valois 1681, 672–76. 11. Wolfgang Seyfarth (1978, xiv) observes that in editing Ammianus he had found that many “vulgate” conjectures were in fact the work of Adrien. 12. A. de Valois 1681, third page, unnumbered, of the praefatio. 13. Burgess’ argument implicitly assumes their composition at a much earlier date. Notes and Discussions 235 Valesiani pars tertia). If in places the chapter headings contain information absent from the text of Ammianus but paralleled in other authors of the late fourth century, the obvious assumption would be that a savant like Adrien knew the texts of Eutropius, Jerome and the Epitome de Caesaribus very well, and that, beyond his direct knowledge of these authors, their information and language may have influenced him indirectly via the abundant scholarly literature both on fourth-century chronology and on Ammianus (much of the latter incorporated into his edition in the form of the commentaries by Lindenbrog and his brother). And this assumption is correct. But it will be worthwhile to say something about the particular interests and textual influences that underlie Adrien’s chapter headings and divisions. One consistent feature of the headings, reflecting one of the great strengths of Henri de Valois’ commentary, and anticipating some of the central concerns of nineteenthand twentieth-century scholarship, is their precision about names and bureaucratic and military ranks. Emperors’ names are usually followed by Augustus or Caesar, and Gallus is twice (14.9, 14.11) referred to as Constantius Gallus, a form not used by Ammianus. 14 The epitome of Book 15 refers very precisely to Ursicinus as Magister equitum per Orientem (15.2) and Silvanus as Magister peditum per Gallias (15.5), though Ammianus did not refer in 15.2 to Ursicinus’ post and in 15.5.2 calls Silvanus pedestris militiae rector; similarly, in place of Ammianus’ circumlocutions, the epitomes give the exact offices held by Valentinian and Jovian at the time when each became emperor (25.5, 26.1). 15 The heading of 27.8 refers to the elder Theodosius as Comes during the British campaign of 367–68: a correct inference, though the text, with a “literary archaism,” refers to him simply as dux. 16 Most of the individuals named in the chapter headings are holders of high offices, such as magistri militum and urban and praetorian prefects, or the leaders of Rome’s enemies. Ammianus’ interest in the high ranks of the military and bureaucracy is accurately reflected in the chapter headings, even if the diversity of his enormous cast list is lost. 17 Adrien also strives for geographical precision, particularly as far as concerns the movements of emperors. In this he is usually commendably accurate, but on at least one occasion he goes astray: the acclamation of Valentinian II is said to take place at Brigetio (30.10), where his father had died six days earlier. Ammianus had last mentioned the name of that city five chapters earlier (30.5.15), and here says simply that Valentinian II was led in castra and acclaimed. We should rather follow 14. Only passing reference is made (14.1.1) to the fact that the emperor Constantius had bestowed his own name on his cousin when he appointed him Caesar. 15. Ammianus is often accused, especially by military historians, of being unclear about civil and military offices. He standardly varies titles (whether for reasons of prose rhythm or to avoid jargon), but any reader with minimal knowledge and interest in administration would understand, for example, that domesticorum ordinis primus (25.5.3) meant primicerius domesticorum. 16. Tomlin 1974, 303 n. 6. The problem is noted in PLRE 1, s.v. Flavius Theodosius 3, and Frakes 2000a, 427. Both from the importance of the British command and from his subsequent promotion to magister equitum it can be inferred that Theodosius was indeed already a comes at this stage (see Tomlin, as well as Birley 2005, 439). Henri de Valois had casually referred to Theodosius as having been comes during his British expedition (ad 28.3.7), as had Adrien in an earlier work (A. de Valois 1646, in the Chronicon, to be discussed below). It seems likely that the brothers calculated his rank from their knowledge of Late Roman government, though it is not impossible that they referred to Theodosius as comes simply because he was often so named to distinguish him from his son the emperor (e.g., Orosius 7.33.7). See also n. 22 below. 17. See Frakes 2000a for a prosopography of 480 contemporaries named in the Res Gestae. 236 Notes and Discussions the explicit statement of the Consularia Constantinopolitana for the year 375, that the younger Valentinian was acclaimed in Aquincum, modern Budapest, forty or so miles away from Brigetio. There are also some inconsistencies worth noting. The epitome of Book 14 (the first surviving book) differs notably in style from those that follow: 1. Galli Caesaris saevitia. 2. Isaurorum incursiones 3. Persarum commentum irritum. 4. Saracenorum irruptiones, et mores. 5. Magnentianorum supplicia. 6. Senatus populique Romani vitia. 7. Galli Caesaris immanitas et saevitia 8. Orientis provinciarum descriptio. 9. De Constantio Gallo Caesare. 10. Pax Alamannis petentibus datur a Constantio Augusto. 11. Constantius Gallus Caesar evocatur a Constantio Augusto et capite truncatur. The brevity—a mere forty-seven words, only two sentences with verbs—is atypical, as is the consequence, that it is impossible to distinguish between the subject matter of 14.1, 14.7, and 14.9. All the subsequent books offer far more detail (for example, Book 15 is epitomized in 126 words, Book 31 in 298). It is unusual in subsequent headings for an urban prefecture to pass unmentioned, as in 14.6. It appears that Adrien quickly changed his mind about the amount of detail his headings would require. A further sign of indecision, perhaps arising from haste, occurs in Book 25: the initial epitome indicates eleven chapters, but the book has only ten, as in modern editions. The error arose in chapter 3, which describes the wounding and death of the emperor Julian. Adrien’s epitome reflects an earlier intention to divide it into two parts: 18 3. Imperator, dum ad repellendos Persas qui undique instabant, omissa lorica temere se proeliis inserit, hasta vulneratur, ac in tabernaculum refertur. 4. Romanis hinc victoribus, inde fusis, Julianus in tentorio jacens circumstantes alloquitur, ac post epotam frigidam moritur. Modern editions follow the lead of Wagner and Erfurdt (1808), who combined the two sentences (. . . in tabernaculum refertur, ubi circumstantes alloquitur . . .) and adjusted the numbers of the following chapters to fit the chapter divisions long established in the text. In some places it is possible to identify more closely external influences (that is to say, outside the text of Ammianus, which was obviously central) on Adrien’s chapter headings. What follows is unlikely to be a comprehensive list of “sources,” but will give a general impression of the technique with which the epitomes were compiled, and incidentally account for those places where they have been observed to contain information absent from the text. First, only one out of the dozen or so earlier editions of Ammianus had chapter divisions and headings, and Adrien had certainly looked at it. The text, published in Lyon in 1591 and reprinted in 1600 “apud Franciscum Le Preux,” has its own distinct epitomes and scheme of chapter divisions. 19 (In these features as in others, this edition is followed by the first English and French translators, Philemon Holland (1609) and Michel de Marolles (1672), though the latter also takes much from Henri de Valois’ 1636 edition.) The “Le Preux” edition had tended to split longer compo18. The fourth chapter would have started at 25.3.15. 19. This volume contains a chronology of the period from Nerva to Valens and a Gnomologia Marcelliniana; its margins are filled out with an idiosyncratic moralizing commentary. Notes and Discussions 237 sitional blocks into several chapters, so that, for example, the digression on the Black Sea was given four chapters, that on Persia and the East seven, and the long account of the magic and treason trials in Rome five: Adrien unites these into single chapters (22.8, 23.6, and 28.1). At other points he splits chapters of the “Le Preux” edition into several parts. Though on occasion the vocabulary of his chapter headings was picked up from his predecessor, the influence is largely a negative one. Looking, for example, at Book 22, one sees how much more fully Adrien’s headings acknowledge the emperor Julian’s movements, from Dacia through Thrace into Constantinople (1–2), and subsequently (9) to Antioch via Nicomedia and Ancyra; the “Le Preux” headings mention only Constantinople and Antioch. The older edition’s description of Julian veers between enthusiastically endorsing the exemplary aspects of Ammianus’ portrayal and a Christian partisanship that sets it at odds with the text it summarizes, as can be seen in the heading to its chapter 10 (which coincides with ours): Antiochiae hyemat, et ibi patientiae, lenitudinis, aequitatis documenta exhibet, sed verae religionis infensus et professus hostis grammaticos et rhetores Christianos a scholis arcet. Here, as elsewhere, Adrien is faithful to the text’s generally understated coverage of Julian’s religious policies, and his heading follows his author’s emphasis that religious affiliation did not bias Julian as a judge (Antiochiae hiemans Julianus jura reddit, nec quemquam propter religionem gravat). But his omission of Julian’s teaching decree, which, briefly and exceptionally, incurs Ammianus’ emphatic criticism (22.10.7), may seem less faithful. Secondly, a number of the chapter headings are taken over from the work that had been the foundation of Adrien’s reputation twenty-five years earlier. The Gesta Veterum Francorum, covering the history of the Franks from the middle of the third century to the deposition of Childeric III, was published in three volumes between 1646 and 1658. At the beginning of the first volume is a Chronicon Rerum Francicarum, or Chronicon Francicum, which lists by year and consuls important events in history that involved the participation of the Franks—something for which Ammianus was a major source. A short excerpt will suffice to show the extent to which Adrien drew on his earlier work (see table 1). 20 The entries from his earlier chronology were not taken over completely unchanged. Ammianus focused in 14.11 on how Constantius persuaded Gallus to come to the West, and in 15.5 on how a secret mission in which he had himself participated managed to dispose of Silvanus: these themes are recognized by Adrien’s adaptations. Some information absent from Ammianus is removed: the fact that the Franks held Cologne for ten months, or that the king of the Chamavi was called Nebiogastes. But the information that Silvanus had reigned for twenty-eight days, recorded in Aurelius Victor, Jerome, and the Epitome de Caesaribus, was given a permanent place in editions of Ammianus, although the author had not chosen to record it. There are two further points at which Burgess’ parallels between the chapter headings and other texts can be explained by the use of this chronicle: the form of words used to describe Valens’ elevation (a. 364: Valentinianus . . . Valentem fratrem in consortionem Imperii recepit ~ 26.4: Valentinianus Valentem fratrem . . . consortem 20. A. de Valois 1646, unpaginated. Adrien presumably found his Chronicon Francicum useful for the edition’s running chronology of years and consuls at the top of each page. My quotations omit the consular dating. 238 Notes and Discussions Table 1. Adrien de Valois’ Self-Imitation Adrien de Valois, Gesta Veterum Francorum, Vol. 1 (Paris, 1646) 354 Constantius Caesar iussu Constantii Augusti capite truncatur. 355 Siluanus, Francus, Magister peditum per Gallias, Coloniae Agrippinae Augustus appellatur, et XXVIII imperii die interficitur. Iulianus a Constantio Augusto Caesar creatur, et praeficitur Galliis VIII idus Novemb. . . . 356 Iulianus Caesar Coloniam a Francis ante X menses captam recipit, et pacem ibidem cum Francorum regibus facit. 359 Iulianus Caesar Saliis et Chamauis, Francicis gentibus, bellum infert: Salios complures in deditionem accipit: cum Chamauorum rege Nebiogaste pacem acceptis obsidibus facit. Adrien de Valois, Edition of Ammianus (Paris, 1681) 14.11 Constantius Gallus Caesar evocatur a Constantio Augusto et capite truncatur. 15.5 Silvanus Francus, Magister peditum per Gallias, Coloniae Augustus appellatur, et XXVIII imperii die per insidias opprimitur . . . 15.8. Julianus Galli frater, a Constantio Augusto fratre patrueli Caesar creatur, ac praeficitur Galliis. 16.3 Julianus Caesar Coloniam a Francis captam recipit, et pacem ibi cum Francorum Regibus facit. 17.8 Julianus Caesar Salios, gentem Francicam, in deditionem accipit: Chamavorum alios caedit, alios capit, reliquis pacem tribuit Imperii constituit), 21 and the location of a Saxon raid specifically in Gaul (a. 370 ~ 28.5). It should be noted that chapter headings for certain other Western events are linguistically indebted to the Chronicon Francicum (a. 360 ~ 20.4, a. 365 ~ 27.1, a. 366 ~ 27.2), but that although most of Adrien’s chronicle entries for the 370s derive from Ammianus, they do not find their way back into the chapter headings of his edition. 22 Thirdly, Burgess’ collection of parallels suggests that Adrien may at three points have used three adjacent passages of Eutropius as a basis for his chapter headings, rather as he used his own earlier work. The first of these is not probative (Eutropius 10.16.1 ~ 24.2), but the other two are both persuasive: Julian was fatally wounded, having thrust himself rashly into battle (dum se inconsultius proeliis inserit, Eutropius 10.16.2 ~ dum . . . omissa lorica temere se proeliis inserit, 25.3); Jovian’s peace was necessary but shameful (exercitu . . . inopia laborante . . . pacem cum Sapore necessariam quidem sed ignobilem fecit, Eutropius 10.17.1 ~ Fame et inopia suorum adductus Jovianus Augustus pacem cum Sapore necessariam sed turpissimam facit, 25.7). In this last case, as Burgess has noted, the heading actually contradicts Ammianus’ belief that Jovian’s peace was unnecessary. 23 21. Burgess (2005) suggests a verbal similarity to Jerome 243e (Valentinianus fratrem Valentem Constantinopoli in communionem regni assumit), a passage that may well have influenced the Gesta Veterum Francorum. Adrien specifies that Valens’ proclamation occurred at the Hebdomon, though Ammianus refers only to a suburbanum (26.4.3): this additional information arises from his own research (see n. 10 above). 22. The Chronicon, though not the text of the Gesta, refers to Theodosius in his campaigns of 367–68 as comes, like the heading of 27.8; cf. n. 16 above. 23. Burgess 2005, 186 n. 54. One Line Short Notes and Discussions 239 A fourth influence (found only once, to my knowledge, and perhaps subconscious) was Henri de Valois’ edition of the church historian Socrates. 24 Adrien gave 22.4 the heading Eunuchos omnes, et tonsores ac coquos Palatio expellit Julianus Augustus, et de Palatinorum spadonum vitiis, ac de corrupta disciplina militari. Ammianus tells a fine story of Julian’s putdown to an overdressed barber, and says that Julian expelled everybody of this sort, along with cooks and the like (22.4.9–10), and he abuses some of the palatini as temple robbers (22.4.3), and attacks the soldiers’ luxury and indiscipline (22.4.6–8). But eunuchs can be found nowhere in the chapter. This is surprising, as we know from Libanius and Socrates that the eunuchs of Constantius’ court were purged, and Ammianus seems likely to have been thinking about eunuchs, whom he generally disliked. (Moreover, the last event described in 22.3 was the execution of Constantius’ sinister eunuch chamberlain, Eusebius). The notes of Henri’s edition of Ammianus point toward the relevant passage of Libanius and also refer to Socrates. But it is his later edition of Socrates that seems to have influenced his younger brother at this point. Socrates (Historia Ecclesiastica 3.1) expands at some length on Julian’s purge of the court, and Henri’s parallel Latin translation includes the words eunuchos vero, tonsores et coquos ejecit ex palatio (translating ejxevbale de; tΩn basileÇwn eu˚nouvcouÍ, koure∂Í, mageÇrouÍ), and he also uses the word expellere for the same action a few lines below. To expand this brief survey of the sources for the chapter headings would be of limited value. Enough is known to draw conclusions about the various parallels with other texts observed by Burgess: 15.5, 26.4, and 28.5 derive from Adrien’s own Gesta Veterum Francorum; 24.2, 25.3, and 25.7 directly from Eutropius; 16.12, 26.2, 26.10, and 30.5 probably show verbal coincidences; 25 at 25.3 and 26.1 the names of offices not given explicitly by Ammianus have probably been inferred from the text, the notes, or external knowledge; likewise at 27.8 (not noted by Burgess); 22.4 (not observed by Burgess) comes from Henri de Valois’ edition of Socrates. Adrien’s division of the text into chapters certainly improves on the “Le Preux” edition. The author’s transitional formulae very often provided a clear guide to where one block of narrative ends and another begins. 26 Timothy Barnes’ recent book offers a definitive analysis of the compositional blocks in which Ammianus wrote, and the congruence between his divisions and Adrien’s is reassuring. 27 It is arguable, however, that occasionally transitional sentences have been misplaced: so in 14.4, the description of Saracen raids in the first sentence belongs with other Eastern events of the year 353, which have been narrated in the previous chapter; the digression on Saracen habits only begins with 14.4.2. The last sentence of 22.8 (48) does not belong with that chapter’s digression on the Black Sea, but fits better with the political 24. H. de Valois 1668; on Socrates, see most recently Urbainczyk 1997. 25. Apud before the site of a battle (16.12) and the pairing Augustus appellatus (26.2) are very common. Adrien likes using the suffix -anus, as at 26.10, for partisans of dead rulers (cf. 14.5 Magnentianorum, 22.3 Constantiani). The phrase ictu sanguinis for Valentinian’s demise is probably just a way of describing a stroke (cf. French coup de sang). 26. See Frakes 2000b, 48 n. 4. For example, the digression on the Saracens ends, hactenus de natione perniciosa; nunc ad textum propositum revertamur (14.4.7). The next chapter opens, Dum haec in oriente aguntur, Arelate hiemem agens Constantius . . . (14.5.1). 27. Barnes 1998, 32–42. 240 Notes and Discussions narrative of the following chapter. 28 The chapter divisions also become less representative of the text in the more episodic and fast-moving final six books.29 To conclude: Adrien’s chapter divisions on the whole reflect Ammianus’ own narrative blocks sensitively and effectively. The arrangement of chapters, and also the sentence divisions (introduced by Wagner and Erfurdt in 1808) have in any case become hallowed through use. 30 The chapter headings are often pithy, and are undeniably helpful as a means of reference. That said, the composition of the headings appears to have been haphazard and magpie-fashion. Adrien did not really hit his stride until after Book 15, and even thereafter, the headings often fail to indicate important episodes in the text (for example, the great tsunami of 365, 26.10.15–19). In places the headings gloss the text with information absent from Ammianus: the twenty-eight-day reign of Silvanus at 15.5, the expulsion of eunuchs from Julian’s court at 22.4, the identification of the Hebdomon as the site of Valens’ proclamation at 26.4, Theodosius’ rank of comes in 27.8, Gaul as the location of a Saxon attack at 28.5. At 25.7 (Jovian’s peace with Persia) the summary puts forward a view to which the historian was opposed. Most significantly of all, at 30.10 the acclamation of Valentinian II is wrongly placed in Brigetio rather than in Aquincum. There is a risk that casual readers may believe that Ammianus says something he does not.31 Editors subsequent to Adrien have been content to print his chapter headings without comment, generally as an epitome at the start of each book, as Adrien had done. The extent to which this has been done on automatic pilot can be exemplified by the fact that when Adrien printed the initial epitome of Book 25 with eleven chapters but the book itself with only ten, both Gronovius (1693) and Ernesti (1773) retained his mistake. The standard modern edition, Seyfarth’s Teubner, also falls beneath its usual high standards in the treatment of Adrien’s chapter headings, printing them at the start of books, but without explaining their authorship in either the apparatus or the introduction. As in other modern editions, some limited changes have been made to their text from Adrien’s original: names that recent scholarship has found to be corrupt or misspelled have rightly been altered, so the prefect of Rome is Viventius not Iuventius (27.3), the Armenian king is Papa not Para (30.1), and the city at which Valentinian II’s proclamation occurred (or rather, as we have seen, did not occur) is Brigetio not Bregitio (30.10). But Seyfarth’s normalization of orthography has missed ditionem at 26.8 and tentant at 31.16, and abbreviations like A. or Aug. for Augustus have not always been expanded. It is understandable that the seventeenth-century abundance of Adrien’s punctuation has been reduced, but Seyfarth often goes beyond minimalism, and in one instance the punctuation is misleading as well as confusing.32 28. Barnes 1998, 34; Frakes 2000b, 48 n. 4. 29. See Barnes 1998, 39–42. 30. As well as introducing the universally accepted division of chapters into numbered sentences, Wagner and Erfurdt’s is the first edition to divide the text within chapters into (unnumbered) paragraphs. In this, modern editions generally follow their lead (see, e.g., Yonge 1862, vii: “Wagner and Erfurdt’s . . . division of chapters into short paragraphs has been followed”). However, the paragraph divisions are not always beyond criticism: see Barnes 1998, 174 n. 30, on 14.6. 31. At 31.16 the heading is likely to mislead readers into believing that there was such a post as Magister militum trans Taurum. 32. For example, the restoration of commas would be helpful to readers at 19.3 and 31.11, where Seyfarth prints Vrsicinus noctu obsidentibus superuenire frustra conatur Sabiniano Magistro militum repugnante and Gratianus Augustus ad patruum Valentem properat ei contra Gothos laturus auxilium. At 25.7 ( fame et inopia suorum adductus Iouianus Augustus pacem necessariam, sed turpissimam facit traditis V regionibus cum Nisibi et Singara) the comma was originally after facit, not necessariam. In these cases Long Notes and Discussions 241 Future editors should consider whether they should maintain the tradition of printing these headings, or omit them altogether, or compose their own. Editors who print them should at least make their origins clear to readers. Gavin Kelly University of Edinburgh Seyfarth, perhaps inappropriately for an early modern text, applies modern German rules of punctuation. For Valentinian’s death at 30.6 (Idem dum legatis Quadorum populares suos purgantibus respondet, iratus ictu sanguinis exstinguitur), there is no such excuse. The punctuation perverts the sense, and the comma should be restored to its original position after iratus. LITERATURE CITED Editions and Translations of Ammianus Marcellinus (in Chronological Order) “Apud Franciscum Le Preux.” 1591. Ammiani Marcellini Rerum sub Impp. Constantio, Iuliano, Ioviano, Valentiniano, et Valente, per XXVI annos gestarum historia: Libris XVIII comprehensa, qui e XXXI hodie supersunt. Cui accesserunt breviaria, notae, Chronologia Marcelliniana, Gnomologia Marcelliniana, index. Lyon. Repr. 1600. Holland, P. 1609. The Roman historie . . . Wherunto is annexed the chronologie . . . together with compendious annotations. . . . London. Lindenbrog, F. 1609. Ammianus Marcellinus. Rerum gestarum qui de XXXI supersunt, libri XVIII ad fidem ms. et veterum codd. recensiti, et observationibus illustrati, ex bibliotheca F. Lindenbrogii. Hamburg. Valois, H. de. 1636. Ammiani Marcellini Rerum gestarum qui de XXXI supersunt, libri XVIII ex ms. codicibus emendati ab Henrico Valesio, et annotationibus illustrati. Adjecta sunt excerpta de gestis Constantini nondum edita. Paris. Marolles, M. de. 1672. Les dix-huit livres qui nous restent des XXXI. de l’histoire qu’ avoit composez Ammian Marcellin (Fragment de l’histoire d ’un autheur inconnu qui estoit Chretien. Touchant quelques actions des empereurs Constantius Chlorus, Constantin le Grand, etc. Remarques de M. H. Valois—La Vie d ’A.M.). Paris. Valois, A. de. 1681. Ammiani Marcellini Rerum gestarum qui de XXXI supersunt libri XVIII ope mss. codicum emendati ab Henrico Valesio et auctioribus adnotationibus illustrati. Necnon excerpta vetera de gestis Constantini et regum Italiae. Editio posterior, cui Hadrianus Valesius, Historiographus Regius, Fr. Lindenbrogii J.C. in eundem historicum ampliores observationes, et collectanea variarum lectionum adjecit; et beneficio codicis Colbertini multis in locis emendavit, notisque explicuit; disceptationem suam de Hebdomo, ac indicem rerum memorabilium subjunxit. Praefixit et praefationem suam, ac vitam Ammiani a Claudio Chiffletio J.C. compositam. Paris. Gronovius, J. 1693. Ammiani Marcellini Rerum gestarum qui de XXXI supersunt, libri XVIII, ope mss. codicum emendati ab Frederico Lindenbrogio et Henrico Hadrianoque Valesiis, cum eorundem integris observationibus et annotationibus, item excerpta vetera de gestis Constantini et regum Italiae. . . . Leiden. Ernesti, A. G. 1773. Ammiani Marcellini Rerum gestarum libri qui supersunt ex recensione Valesio-Gronoviana. Leipzig. Wagner, J. A., and C. G. A. Erfurdt. 1808. Ammiani Marcellini quae supersunt, cum notis integris F. Lindenbrogii, Henr. et Hadr. Valesiorum et I. Gronovii, quibus T. Reinesii quasdam et suas adiecit I.A. Wagner. Ed. absolvit C. G. A. Erfurdt. Leipzig. Repr. Hildesheim, 1975. Yonge, C. D. 1862. The Roman History of Ammianus Marcellinus during the Reigns of the Emperors Constantius, Julian, Jovianus, Valentinian, and Valens. London. Clark, C. U. 1910–15. Ammiani Marcellini Rerum gestarum libri qui supersunt. Recensuit rhythmiceque distinxit C. U. Clark, adiuvantibus L. Traube et G. Heraeo. Berlin. 242 Notes and Discussions Rolfe, J. C. 1935–39. Ammianus Marcellinus. Loeb Classical Library. 3 vols. Cambridge, Mass. Galletier, E., with J. Fontaine. 1968. Ammien Marcellin, Histoire Tome I: Livres XIV–XVI. Paris. Seyfarth, W. 1968–71. Ammianus Marcellinus: Römische Geschichte. Lateinisch und deutsch und mit einem Kommentar. Berlin (4 vols). , with L. Jacob-Karau and I. Ulmann. 1978. Ammiani Marcellini Rerum gestarum libri qui supersunt. Leipzig. Repr. Stuttgart and Leipzig, 1999. Hamilton, W. 1986. Ammianus Marcellinus, The Later Roman Empire (a.d. 354–378). Selected and Translated by Walter Hamilton; with an Introduction and Notes by Andrew Wallace-Hadrill. Harmondsworth. Other Literature Cited Barnes, T. D. 1970. The Lost Kaisergeschichte and the Latin Historical Tradition. BHAC 1968–69: 13–43. . 1978. The Sources of the “Historia Augusta.” Brussels. . 1998. Ammianus Marcellinus and the Representation of Historical Reality. Ithaca, N.Y. Bird, H. W. 1973. Further Observations on the Dating of Enmann’s Kaisergeschichte. CQ 23: 375–77. Birley, A. R. 2005. The Roman Government of Britain. Oxford. Broszinski, H., and H. C. Teitler. 1990. Einige neuerdings entdeckten Fragmente der Hersfelder Handschrift des Ammianus Marcellinus. Mnemosyne 43: 408–23. Burgess, R. W. 1993. Principes cum Tyrannis: Two Studies on the Kaisergeschichte and Its Tradition. CQ 43: 491–500. . 1995. On the Date of the Kaisergeschichte. CP 90: 111–28. . 2005. 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. CP 100: 166–92. Clark, C. U. 1904. The Text Tradition of Ammianus Marcellinus. New Haven, Conn. Dionisotti, A. C. 1997. Les chapitres entre l’historiographie et le roman. In Titres et articulations du texte dans les œuvres antiques, ed. J.-C. Fredouille et al., 529–44. Paris. Enmann, A. 1884. Eine verlorene Geschichte der römischen Kaiser und das Buch “De viris illustribus urbis Romae”: Quellenstudien. Philologus Suppl. 4.3, 335–501. Frakes, R. M. 2000a. Ammianus Marcellinus and His Intended Audience. In Studies in Latin Literature and Roman History 10, ed. C. Deroux, 392–442. Collection Latomus 254. Brussels. . 2000b. Some Thoughts on the Length of the Lost Books of Ammianus Marcellinus. Ancient World 31: 48–59. Paschoud, F. 2006. Chronique d’historiographie tardive. AnTard 14: 325–44. Reynolds, L. D. 1983. Ammianus Marcellinus. In Texts and Transmission, ed. L. D. Reynolds 6–8. Oxford. Robinson, R. P. 1936. The Hersfeldensis and the Fuldensis of Ammianus Marcellinus. University of Missouri Studies 11: 118–40. Schröder, B.-J. 1999. Titel und Text: Zur Entwicklung lateinischer Gedichtüberschriften, mit Untersuchungen zu lateinischen Buchtiteln, Inhaltsverzeichnissen und anderen Gliederungsmitteln. Berlin. Tomlin, R. S. O. 1974. The Date of the Barbarian Conspiracy. Britannia 4: 303–9. Urbainczyk, T. 1997. Socrates of Constantinople: Historian of Church and State. Ann Arbor, Mich. Valois, A. de. 1646–58. Gesta Veterum Francorum. 3 vols. Paris. . 1677. Hadriani Valesii Historiographi Regii de Vita Henrici Valesii Historiographi Regii Liber. Paris. Repr. Leipzig, 1680. Valois, H. de. 1668. Socratis Scholastici et Hermeiae Sozomeni Historia Ecclesiastica. Henricus Valesius Graecum textum collatis mss. codicibus emendavit, Latine vertit, et annotationibus illustravit. Paris.