¹ Gothofredus' commentary was not accessible to me when I began to write this paper, and I am deeply endebted to Dr John Martindale (Cambridge), author of the Prosopography of the Later Roman Empire (hereafter PLRE), who transcribed the relevant passage for me and added valuable comments. In the meantime I have seen the edition of Gothofredus which appeared at Mantua in 1748, where the passage will be found on p. 173.

² On this manuscript see Paschoud, F., Zosime, Histoire nouvelle, 1 (Paris, 1971), p. lxxvii ff.Google Scholar

³ See Haury's text of Procopius, Bell. Goth., 5. 8. 4, 7. 6. 5, 18. 20 ff., 28. 7; Agathias, Hist. II. 1, ed. Dindorf, p. 179 init., ed. R. Keydell, p. 40. 17. Keydell reports some minor manuscript variations, but none of any significance for the present discussion.

⁴ See Henry, R., Photius, bibliotheque, Collection byzantine, l (Paris, 1959), p. 171. On Olympiodorus see now B. Baldwin, ‘Olympiodorus of Thebes’, L'antiquite classique 49 (1980), 212–31, with bibliography in the footnotes.Google Scholar

⁵ See Boissevain's edition of Dio, vol. 1, p. 249 n. 10.

⁶ See the Loeb edition, vol. II, p. 442.

⁷ O. Seeck, Untergang, v, p. 598 n. on p. 410. 21, followed by C. E. Stevens,‘Marcus, Gratian, Constantine’, Athenaeum 35 (1957), 316–47, at 330, and Demougeot, E., De I'unitè ⋯ la division de 1'empire Romain (Paris, 1951), p. 457 n. 90Google Scholar, but not by Paschoud, F., P.-W., XA, 824, or by John Matthews, Western Aristocracies and Imperial Court, A.D. 364–425 (Oxford, 1975), p. 298n.Google Scholar Seeck's reason for making his suggestion is that Zosimus, vi. 9. 1 and 12. 1, reports two expeditions to Africa, and, according to Seeck, the second is a doublet of the first. This seems inadequate. See p. 451 below.

⁸ The matter is discussed in Britannia 8 (1977), 303–18, esp. 310 ff.

⁹ See the references in PLRE II, p. 539 f.

¹⁰ Zosimus 6. 7. 5 f., 8. 3 and 12; Sozomen, Hist. Eccles. 9. 8. 3 f., 7 f., who both drew on Olympiodorus. See p. 451 below.

¹¹ Gildas, De Excidio 18, p. 34. 19 ff., Mommsen; p. 94, Winterbottom. Freeman, E. A., Western Europe in the Fifth Century (London, 1904), p. 152 f., saw something of the connexion between Zosimus and Gildas here.Google Scholar

¹² Students of the Roman army in Britain should note for what it is worth - and it is the tendentious remark of a slippery diplomat, reported by an unreliable historian - Jovius' statement implying that there were still Roman troops in Britain even when Constantine had crossed to Gaul: Zosimus 6. 1.2.

¹³ Olympiodorus, frg. 12. That Zosimus is mistaken here was pointed out by Freeman, op. cit. 45 n.^*, and by Baynes, Norman H., Byzantine Studies and Other Essays (London, 1955), p. 339. These two passages ought to have been cited in Britannia 8 (1977), 304 n. 1.Google Scholar

¹⁴ It is not clear to me why Johnson, Stephen, Later Roman Britain (London, 1980), p. 105 f., puts them in 408. If that were correct, it would follow that Zosimus has omitted from his Rückblick all the events that occurred in 409.Google Scholar

¹⁵ Bury, J. B., History of the Later Roman Empire I (London, 1923), p. 183 n. 2.Google Scholar

¹⁶ Sozomen, Hist. Eccles. 9. 9. 3 f., cf. Philostorgius, Hist. Eccles. 12. 3 (p. 142, ed. Bidez), Zosimus 6. 13. 2. Olympiodorus admired the courage of Sarus: see his fir. 3 and 17.

¹⁷ At the beginning of Chapter thirteen Zosimus makes an abortive attempt to leave the affairs of Italy and to narrate events in Gaul, where Constantine elevated his son Constans from the rank of Caesar to that of Augustus. But after no more than three lines of the printed text he abruptly, arbitrarily, and characteristically grows weary of Gaul and returns to Italy, Alaric, and Sarus. But no doubt in his own incompetent way he was following Olympiodorus, for it was Olympiodorus' practice to add to his account of Italian affairs digressions on events in Gaul and Spain: see J. F. Matthews, ‘Olympiodorus of Thebes and the History of the West (A.D. 407–425)’, JRS 60 (1970), 79–97, at 82, 87.

¹⁸ So, e.g., Paschoud, F., Cinq études sur Zosime (Paris, 1975), p. 180n. 5;cf. J. F. Matthews, art. cit., 87.Google Scholar

¹⁹ On TertuUus see PLRE II, p. 1059, s.v. ‘TertuUus i’; cf. p. 1149, s.v. ‘Varanes i’.

²⁰ Stein, E., Histoire du bas-empire I (Paris, 1959), p. 258Google Scholar, dates the elevation of Attalus to ‘novembre 409 environ’, though I would prefer to follow L. Schmidt, Geschichte der deutschen Stdmme: die Ostgermanen (reprinted Munich, 1969), p. 445, ‘Ende 409’, or Demougeot, op. cit., 449, ‘sans doute en decembre 409’.

²¹ Zosimus 6. 7. 5–6.

²² idem 6. 8. On the Eastern troops who came to help Honorius on this occasion see Jones, A. H. M., Later Roman Empire, A.D. 284–602 (Blackwell, Oxford, 1964), p. 682.Google Scholar

²³ Zosimus 6. 9. 2.

²⁴ idem 6. 9. 3.

²⁵ Cod. Theodos. 9. 28. 6.

²⁶ Bury op. cit. I,183 n. 2, following Schmidt, op. cit., 446 n. 4, points out that in Cod. Theodos. 7.16.2 (of 24 April), the words tyrannicifuroris et barbaricae feritatis refer to Attalus and Alaric. For the reason stated in the text I would prefer to follow PLRE II, p. 181, s.v. ‘Priscus Attalus 2’, in dating the deposition to ‘summer 410’, or, better still, Schmidt, op. cit., 447, who dates it to July 410. Stein, op. cit I,259, also dates the deposition to‘ juillet 410 environ’: cf. Demougeot op. cit., 462. But deserters from Attalus were coming over to Honorius as early as February: Cod. Theodos. 9. 38. 11.

²⁷ Cod. Theodos. 9. 38. 12. On this Palladius see PLRE II, p. 819. s.v. ‘Palladius 2’.

²⁸ Zosimus 6. 8. 1. Notice Honorius' changes of mood in these weeks. From terror he passes to a rather over-confident plan to attack Alaric (Zosimus 6. 8. 3), but fails - perhaps wisely - to carry it out. (I say ‘perhaps wisely’ because in 409 he had sent 6,000 crack troops from Dalmatia against Alaric with the result that they were all but annihilated.) We next find that when money reaches him from Heraclian he is in a state of euphoria (ibid. 10. 2). We have no such psychological touches in our descriptions (such as they are) of the other fifth-century emperors; and they make us realize more than ever the catastrophe we have suffered in losing the work of Olympiodorus. For his powers of characterization note especially his pen-portrait of Constantius (later Constantius III) in frg. 23.

²⁹ Olympiodorus, frg. 13.

³⁰ idem, frg. 14, cf. Sozomen, Hist. Eccles. 9. 12. 5.

³¹ For a convenient method of converting Olympiad dates into the corresponding years of the Christian era see C. Courtois, Byzantion 21 (1951), 31. The magic formula for this transmutation is: ( x - 1)4-(776–1), where x is the number of the Olympiad.

³² See Mommsen, Chron. Min., I, 618 ad fin.

³³ A good start in coming to grips with this work has been made by M. Miller, ‘The Last British Entry in the “Gallic Chronicle”', Britannia 9 (1978), 315–18, and P. J. Casey, ‘Magnus Maximus in Britain: A Re-Appraisal’, The End of Roman Britain, ed. P.J. Casey, BAR British Series No. 71 (1979), 66–79.

³⁴ Chron. Gall. a. CCCCLII, 63 (I, 654). For the date of the entry of the invaders into Spain see Hydatius, 42 (ii, 17).

³⁵ Chron. Gall. a. CCCCLII, 64.

³⁶ For an attempt to explain how the chronicler's mistake about this John, if it is a mistake, might have arisen, see PLRE II, p. 594, s.v. ‘Joannes 5’.

³⁷ Zosimus 6. 5. 2.1 cannot think of any reason that would justify Stephen Johnson, op. cit. (in n. 14), 109 fin., in asserting that ‘Britain declared her independence from Rome in 406’ or that ‘the necessity of dissociating themselves from Rome's rule may have been as clear for the Britons in 406 as Zosimus says it was in 408 (sic), and may have resulted therefore in the choice of Marcus as British Emperor. ‘The only indication that we have of the Britons’ motives in 406–7 is given by Sozomen, HE 9. 11. 2, of which a translation will be found in Britannia 8 (1977), 318.

³⁸ Zosimus 6. 5. 3.

³⁹ Britannia 8 (1977), 310 f.

⁴⁰ Freeman, op. cit. (in n. 11 above), 148.

⁴¹ Rutilius Namatianus, De Reditu Suo I, 213–6. The word oras there does not necessarily imply that only the coastal regions of Armorica were being suppressed by Exuperantius. For the date of the De Reditu Suo see Alan Cameron,‘Rutilius Namatianus, St Augustine, and the Date of the De Reditu’, JRS 57 (1967), 31–9.

⁴² Jones, op. cit., 187, cf. 1023. For a possibly even more exotic opinion see Momigliano, A., Alien Wisdom: The Limits of Hellenization (Cambridge, 1975), p. 73, who, appealing to Zosimus 6. 5. 3 remarks that ‘Celtic self-awareness may be traced in the popular rebellions of the fifth century, especially in Armorica’.CrossRefGoogle Scholar

⁴³ J. P. C. Kent, ‘The End of Roman Britain: The Literary and Numismatic Evidence Reviewed’, apud P. J. Casey, op. cit., (in n. 33 above), 15–27, at 18, writes, ‘They were imitated by the Armoricans and others. So there is no question here of Bagaudae, but of magnates’. If that word ‘so’ means ‘therefore’ - and even if it does not - I find the connexion of thought in this quotation incomprehensible.

⁴⁴ Britannia 10 (1979), 208 f.

⁴⁵ Nash-Williams, V. E., The Early Christian Monuments of Wales (Cardiff, 1950), p. 107, no. 139.Google Scholar

⁴⁶ Rutilius Namatianus, 213–16.

⁴⁷ Stephen Johnson, ‘Channel Commands in the Notitia’, apud R. Goodburn and P. Bartholomew, Aspects of the Notitia Dignitatum, BAR Supplementary Series 15 (1976), 81–102, at 82, citing Amm. Marc, 27. 8. 5, 28. 2. 12, 5, 7.

⁴⁸ See J. G. F. Hind, ‘Litus Saxonicum - The Meaning of the “Saxon Shore”, apud W. S. Hanson and L. J. F. Keppie (eds.), Roman Frontier Studies, BAR International Series 71 (1980), 317–24.

⁴⁹ Orosius, Hist. 7. 40. 4.

⁵⁰ Prosper, Chron. 1301 (Chron. Min., I, 472), referring to the year 429.

⁵¹ ibid. 1307.

⁵² Britannia 10(1979), 217 f. I cannot account for the fact that Salvian never mentions Britain. For a suggestion see R. P. C. Hanson,‘The Reaction of the Church to the Collapse of the Western Roman Empire in the Fifth Century’, Vigiliae Christianae 26 (1972), 272–87, at 274, cf. 282.

⁵³ Jerome, Ep. 133; cf. Procopius, Bell. Vandal., 3. 2. 31.

⁵⁴ 54 On Maximus see PLRE n, 744 f., s.n. ‘Maximus 4’. It is not quite certain that he wasg Gerontius' son.

⁵⁵ St Augustine does not appear to have been interested in the fate of the British provinces. More credit to Olympiodorus, far away in Egypt, for finding out the facts and reporting them.

⁵⁶ The text will be found in Mommsen, Chron. Min., I, 629 f. See Mommsen's remarks ibid. 617.

⁵⁷ Procopius, loc. cit.

⁵⁸ cf. Britannia 8 (1977), 306. For a new and interesting interpretation of Honorius' letters, though one which I cannot accept, see M. Miller, ‘Stilicho's Pictish War’, Britannia 6 (1975), 141–5, at 145 n. 20, who goes on to ask (ibid n. 22), why Bede adds tempore autumni to Gildas' narrative. For the answer to this question see Freeman, op. cit. (in n. 11 above), 151 n.