¹⁵² In their original article the following entries are claimed to be Carolingian interpolations on the basis of no evidence other than the fact that they are thought by Jones and Casey to be in error: 1, 10, 24, 32, 39, 42, 54, 58, 83, 94, 122, 137. The Olympiads and years of Abraham were also claimed to be interpolations for the same reason. The following entries are claimed to have been shifted or redated by scribal corruption on the basis of no evidence other than the fact that they are thought by Jones and Casey to be in error: 30, 31, 53, 82, 85, 88, 89, 90, 91, 92, 93, 129 (cf. Britannia xix (1988), 371 n. 6; 373; 375-9). There is no reason for suspecting any of these entries unless one assumes that the chronicler could not have made such ‘errors’.Google Scholar

¹⁵³ For instance, without inerrancy they have no grounds for redating entry 53 on Arcadius' death to a position between entries 56 and 57 (two years later), the cornerstone of Jones and Casey's acceptance of the chronology of entry 62 on the Saxons. This leaves the chronicler open to charges of gross error at a key point in the Chronicle for an event much better known than the Saxon invasions (cf. rebuttal, pp. 212-13; Britannia xix (1988), 376–7).Google Scholar

¹⁵⁴ cf. Britannia xix (1988), 377–8. The importance of this entry was also stressed by John Casey at an Oxford seminar in December 1984, which I attended.Google Scholar

¹⁵⁵ My attack on Jones and Casey's Latin (‘Chronica Gallica a CCCCLIP’) was petty and uncalled for. But their rebuttal (p. 213) shows that they still do not understand the point that their ‘a’ (= preposition) is different from Mommsen's ‘a.’ (=a(nni)/a(nno)).

¹⁵⁶ They state that I ‘fail to undermine the validity of the authors’ view that these regnal years are computed in relation to imperial seniority’ (p. 215), and that Theodosius created two separate states with Arcadius and Honorius as equal emperors (cf. Britannia xix (1988), 372, and rebuttal, p. 214 n. 37).Google Scholar

¹⁵⁷ See especially T.D. Barnes, The New Empire of Diocletian and Constamine (1982), 3 n. 1; 24 n. 12.

¹⁵⁸ The unimportance of 395 has been realized since the beginning of this century, cf. Palanque, Jean-Rémy, ‘Collégialité et partages dans l'empire romain aux IV^e et V^c siècles’, REA xlvi (1944), 47-64, 280–98, esp. 285-6, and n. 7. I cannot find any modern work that would support Jones and Casey's claims. The passage Jones and Casey misquote (p. 214 n. 37) from a panegyric delivered to Honorius by Claudian in Milan in 398 is a piece of propaganda and is taken completely out of its political context.CrossRefGoogle Scholar

¹⁵⁹ In their original paper they stressed that this method mirrored ‘strict imperial protocol’ (Britannia xix (1988), 370-2, 374)Google Scholar. If their odd calculation were true, then the Chronicle would even start out incorrectly, since §2 on the accession of Theodosius, placed in Gratian I (‘Year One’), took place on 19 January 379, which is not between Aug. and Dec. 378. By Jones and Casey's calculation it ought to be under Gratian II. Another scribal transposition? Notice that elsewhere they can nevertheless discount partial years as regnal years when it does not fit their hypothesis. Here they count the period 24 August to 31 December 378 as ‘Year One’ of Gratian, but they do not count the period 15 August to 31 December 423 as ‘Year One’ of Theodosius II (Britannia xix (1988), 370-1, 374, and rebuttal, pp. 214 and 215). Part of the problem is that they think that Honorius died in October not August (cf. PLRE I: 442). This is the type of ‘insignificant’ error to which I referred above; their original paper is full of such errors.Google Scholar

¹⁶⁰ cf. Festy, Michel, ‘Puissances tribuniciennes et salutations impériales dans la titulature des empereurs romains de Dioclétien à Gratien’, R1DA xxix (1982), 193–234; Barnes, op. cit. (note 157), 25–7; Dietmar Kienast, Römische Kaisertabelle (1990), 37-9, 329Google Scholar; and Burgess, R.W., NC cxlviii (1988), 77–96. Regnal year calculations were slightly different in Egypt, however; cf. Barnes, op. cit. (note 157), 28–9.Google Scholar

¹⁶¹ See now, for instance, Muhlberger, Steve, The Fifth-Century Chroniclers: Prosper, Hydatius, and the Gallic Chronicler of 452 (1990), esp. 2-6, 152, and 146–60; John Drinkwater and Hugh Elton (eds), Fifth-Century Gaul: A Crisis of Identity? (1992), 5–37Google Scholar; Croke, Brian, The Chronicle of Marcellinus Comes in its Contemporary and Historiographical Context (unpub. D. Phil, thesis, Oxford, 1978)Google Scholar; Burgess, R.W., Hydatius: A Late Roman Chronicler in Post-Roman Spain (unpub. D. Phil, thesis, Oxford, 1989); Elizabeth Jeffreys with Brian Croke and Roger Scott (eds), Studies in John Malalas (1990), esp. III-66; and Alden A. Mosshammer, The Chronicle ofEusebius and Greek Chronographic Tradition (1979). Michael Klaassen of the University of Pennsylvania is working on a study of Cassiodorus' chronicle.Google Scholar

¹⁶² For a comparison of the Narratio and 452 in the reign of Honorius, see Muhlberger, Steve in Britannia xiv (1983), 29–30, and his book cited above (note 161), 152-5.Google Scholar

¹⁶³ This hypothesis can certainly be challenged, but not in the way Jones and Casey attempt it.

¹⁶⁴ cf. Stephen Johnson, Later Roman Britain (1980), 95-107.

¹⁶⁵ Jones and Casey reiterate their claim that the death of Arcadius is the only chronological error between 406 and 413 (p. 212). Yet Caelestinus became Bishop of Rome in 422, not 406 (&54); Nestorius became Bishop of Constantinople in 428, not 408 (&58); the Sueves did not occupy ‘Hispaniarum partem maximam’ until the 440s, not in 410 (&64); and entry 63 is so vague that it would be hard to get it wrong if it were put almost anywhere in this block. Jones and Casey stop at 413 because entry 73, the cession of Aquitania to the Goths in 418, and entry 75, Heraclian's attempted usurpation of 413, are dated to 414 (their redating).