¹ Courtois, Christian, ‘Auteurs et scribes: Remarques sur la chronique d'Hydace’, Byzantion xxi (1951), 23–54.Google Scholar See now Chapter Three of ray forthcoming critical edition of Hydatius which exposes the errors of Courtois' approach to the text. An attempt similar to Courtois' was made by Eduard Schwartz with regard to the Chronici canones of Eusebius; on this, see Mosshammer, Alden A., The Chronicle of Eusebius and Greek Chronographie Tradition (1979), 54–6.Google Scholar Historians of the earlier periods of Greek and Roman history will also be familiar with this approach but on a much smaller scale; for a recent example, see Carawan, Edwin M., TAPA cxviii (1988), 211Google Scholar, who repeats the rule for texts I quote below.

² ‘The Gallic Chronicle Restored: A Chronology for the Anglo-Saxon Invasions and the End of Roman Britain’, Britannia xix (1988), 367–98.

³ Note that on p. 367 Jones and Casey cite the name of the chronicle incorrectly: ‘Chronka Gallica a CCCCLII’ (i.e. the ‘Gallic Chronicle from 452’). This is only one instance of a number of difficulties with Latin; cf. for example, p. 371, where ‘diem obiit’, which just means ‘died’, is translated as ‘died on his dies imperii’, and p. 397, where ‘imperii Romanipotentiam deiecit’ is translated as ‘overthrew the power of the Roman empire', i.e. caused the fall of the Roman empire.

⁴ L is the British Library Additional manuscript 16 974, ff. 109r-110v; B is Bamberg Staatsbibliothek Patr. 62 (olim E III 18), ff. 53V-55V; M is Munich University 6, ff. 38r-40v; S refers to the two dozen or so manuscripts of Sigebert of Gembloux's Chronographia which are based on the editio Aquicinensis of 1112/3, since these have a medial L/BM version of the Chronicle appended at the beginning. For my critical edition I have used LBM and three of the earliest twelfth-century manuscripts of the S tradition: Cambrai, Bibliothèque Municipale Classée 965, ff. 52v-55r (Bethmann B1); Douai, Bibliothèque Municipale 799, ff. 54r-56r (Bethmann B5); and Paris. Bibliothèque Nationale 1793, ff. 92v-94r (Bethmann C3).

⁵ Chron. min. I (= MGH: AA IX): 620, 623, 625.

⁶ The authors also claim that L is made up of ‘single parchment sheets’ (p. 367). Medieval codices were never bound in this way and L is, in fact, bound in qualerniones, that is, quires of four parchment sheets, folded in the middle to make eight folia or sixteen pages. The section which contains the chronicles (57r-113v), however, is a bit unusual in that the first quire (ff. 57–63) is a folium short and another (ff. 96–7) is just a bifolium (for a total of 57 folia).

⁷ This statement makes ‘Appendix II: Misplacement of Abraham Years and Olympiads in the G.C. 452’ (p. 398) appear superfluous, but it would seem that the ‘errors’ presented here are the authors' sole proof that the years of Abraham and the Olympiads were the product of ‘the interpolating editor’ and not the original scribe of L.

⁸ One can compare Mommsen's treatment of the Chronicle of 511 – with its huge empty spaces and only a few entries per page – to the cramped epitome of its unique manuscript where there are no spaces or gaps; the regnal years appear within the text, not in the margin; and none of Mommsen's numerous bracketed regnal years appears at all.

⁹ cf. the commonly believed but fallacious statement under ‘2.’ on p. 398 that the last regnal. year of one emperor and the first of his successor were fractions of but a single calendar year and so should be ‘doubled up’ when counting total calendar years. Chroniclers based their regnal year figures on the full and partial years of each emperor's reign, rounded either up or down to give just full years. This method will work if done carefully, but no other chronicler I have examined is as careful or as accurate as Eusebius and Jerome, who, over the last 421 calendar years of the Chronici canones (from the death of Julius Caesar), report 420 regnal years (though the missing year (306) is actually accounted for as a ‘year of persecution’ [228^f, Helm]) with no ‘doubling up’.

¹⁰ Most other Western sources also give Gratian six years and Theodosius eleven, for a total of seventeen years (counting 395), including the Gallic chronicler's two major sources, the text also used by the compiler of the Narratio (see below, n. 14) and Rufinus' Historia ecclesiastica XI.34, but some do correctly give him sixteen years (cf., e.g., the Gallic Chronicle of 511). There is no question of ‘inclusive reckoning’ or anything of that nature. I have studied over a score of regnal year lists and chronicles, and every one counts regnal years as I have described.

¹¹ See also note 6, pp. 371–2. This total of six (for Gratian) and twelve (for Theodosius), would, of course, then give Theodosius eighteen years in total (not seventeen as the chronicler states), but strangely this does not seem to bother the authors.

¹² As can be seen from the Appendix below, the authors' ‘restoration’ of the last years of Honorius (p. 373) shows that they do not understand the alterations made by the chronicler to fit the supposed twenty eight years of Honorius' reign from 396 to 424 into thirty two regnal years.

¹³ On the methods chroniclers used for determining and presenting regnal years, see Burgess, R.W., CP 82 (1987), 340–5Google Scholar , and my forthcoming historiographical study of the Chronicle of Hydatius.

¹⁴ For this text, see Chron. min. I: 617 and 629–30, and my forthcoming edition of the two Gallic Chronicles. One of the manuscripts which contains the Narratio, Biblioteca de la Universidad Complutense de Madrid 134, was copied from a now-lost Visigothic exemplar which can be dated from surviving folia to the eighth or ninth century; this predates the Gallic L by about one hundred years. On this manuscript, see Chapter Two of my forthcoming critical edition of the Chronicle of Hydatius.

¹⁵ No source I have found other than the Narratio gives Honorius a total of thirty two years. I am inclined to believe that this figure arose from a correction made to the total by the compiler of the original source of the Narratio and the Gallic Chronicle. Most sources give Honorius and Arcadius twelve or thirteen years together, and Honorius fifteen after Arcadius' death, for a total of twenty seven or twenty eight years for Honorius. However, it is obvious from both Gallic Chronicles that this original source had a reference to Honorius' tricennalia of 422 (452: §89; 511: §575). Although this celebration was actually calculated from Honorius' dies imperii on 23 Jan. 393, the original compiler assumed that it was counted from the death of Theodosius (Hydatius does the same thing), yet in his source Honorius was only credited with twenty seven years (twelve with Arcadius and fifteen with Theodosius). Thus, assuming a simple scribal error, the compiler changed the ‘XV’ of his source to ‘XX’ (as it now appears in the Narratio) in order to get a thirtieth year for the celebration, in the process crediting Honorius with thirty two not twenty seven regnal years. As can be seen in both Chronicles (and Hydatius) the tricennalia is dated to Honorius' thirtieth year from the death of Theodosius, not from 393.

¹⁶ The authors have mistakenly used the terms junior and senior Augustus as if they were contemporary titles (Iunior and Senior Augustus).

¹⁷ For the vota, see Burgess, R.W., Numismatic Chronicle cxlviii (1988), 77–96.Google Scholar There are also coins of Theodosius II which note his regnal years (ibid., 86–7) and a number of imperial letters where the emperors list their regnal years (for two examples, see Burgess, R.W., ZPE 65 (1986), 218, n. 32).Google Scholar

¹⁸ The paragraph between ‘FINIT CRhONICA EUSEBII’ (sic) and ‘hucusque Hieronimus quae secuuntur Prosper digessit, which the authors claim is ‘an interpolated bridge between two originally independent works … explaining that the preceding source was continued by Jerome and falsely attributing the next source’ (p. 376), is actually a version of Jerome's supputatio from the end of his chronicle, and far from explaining anything about sources, it simply lists the number of years between key Biblical and historical events. It should also be noted here that the Gallic Chronicle was never ‘an independent work,’ but, in fact, always followed Jerome.

¹⁹ The reader is here advised to consult Chron. min. I: 660. Part of the initial plausibility of the authors' arguments derives from the fact that they never explain the actual set-up of the text at this point, a lapse which makes this section highly misleading for those unfamiliar with the text.

²⁰ This is an example of homoearchon since the scribe's eye has gone from “B-' ellum” to “B-'urgundionum’.

²¹ It is seven lines lower in B, five in L. Judging from the serious compression attempted here by the scribe of L, the archetype is more likely to have had seven than five.

²² Chron. min. II: 62 and I: 298, and Ho Peng Yoke, ‘Ancient and Mediaeval Observations of Comets and Novae in Chinese Sources’, Vistas in Astronomy 5 (1962), no. 178, p. 160, dated 7 (or 22) August to 17 September. Philostorgius, HE X.9 (Die Griechischen Christlichen Schriftsteller, (1981), 129–30) describes the appearance of this comet in great detail, but associates it with Theodosius' stay in Rome of 13 June to 30 August 389, though this may be the fault of his epitomator, Photius.

²³ Chron. min. I: 300, and §65 (of my new edition; 73 in Mommsen's). This event, dated 25 July, is apparently not the comet of 17 February, 419 (Ho, op. cit. (note 22), no. 192, p. 162).

²⁴ Structurally, it is important for this entry to mention the Saxons since it provides a focus of chronology and pathos for the first part of entry 126, the final Saxon conquest (see below).

²⁵ The Narratio states ‘Galliae Hispaniaeque a barbarti nationibus Wandalis Suebis Alaras excisae funditusque deletae sunt’.

²⁶ Jones and Casey appear to accept the first restriction (‘To contemporary observers in Gaul, some significant portion of Britain passed into Saxon control in A.D. 441’, p. 396), but they obviously do not understand that this need not have had any bearing on what was actually going on in Britain at the time, since they then immediately go on to state as an historical fact what they had just admitted was merely a ‘Gallic observation’ (‘Within the span of a single generation, A.D. 410–441, the Saxon invaders moved from defeat to victory and the rule of a significant portion of Britain’, p. 397). As for the actual period of the Saxon conquest, since Bede's dates are of no value, coming as they do from Gildas, we are left to fall back on Nennius 66 and archaeology.

²⁷ This appendix follows the corrected chronology of my forthcoming edition in which nine entries have been redated (28, 50, 77, 81, 88, 102, 106, 107, 126). It is not meant to be an authoritative list of accepted dates and events, simply what I felt could be dated reasonably from independent sources.