¹ The ensuing discussion of Bede and the Chronicle is an annotated text of the second of two O'Donnell Lectures, delivered in May 1982 and published by kind permission of the University of Oxford. The first lecture is published as ‘Gildas and the Anglo-Saxons’, Cambridge Med. Celtic Stud. 6 (Winter 1983), 1–30Google Scholar. I am indebted to the comments and advice of Crook, J. A., Dumville, D. N., Keynes, S. D., Lapidge, M., Page, R. I. and Thompson, E. A.. References to the De Excidio Britanniae (hereafter DEB) are to Gildas: The Ruin of Britain and Other Works, ed. and trans. Winterbottom, Michael (Chichester, 1978)Google Scholar. Bede's Historia Ecclesiastica (hereafter HE) is generally quoted from Bede's Ecclesiastical History of the English People, ed. and trans. Colgrave, Bertram and Mynors, R. A. B. (Oxford, 1969)Google Scholar, but reference is also made to the notes in Venerabilis Baedae Opera Historica, ed. Plummer, Charles (Oxford, 1896)Google Scholar (hereafter BOH). The Anglo-Saxon Chronicle is generally quoted according to the translation in English Historical Documents, c. 500–1042, ed. Whitelock, Dorothy, 2nd ed. (London, 1979), pp. 145–261.Google Scholar

² Lappenberg's most lasting critical contribution, the supposed patterning of events in the Chronicle at intervals of four and eight years, has finally been disposed of by Harrison, Kenneth (The Framework of Anglo-Saxon History to A.D. 900 (Cambridge, 1976), pp. 147–50)Google Scholar, whose argument was anticipated by Guest, Edwin (Origines Celticae (London, 1883) 11, 161–2).Google Scholar

³ Kemble, John Mitchell, The Saxons in England (London, 1849) 1, 16.Google Scholar

⁴ Origines Celticae 11, 147. (This volume reprints this and allied papers from ArchJ.) The following quotations are taken from 1, xix–xxi, where ‘E.V.’ also records Kemble's admiration for Guest's ‘intimate acquaintance with Welsh language and literature’: ‘In other things I am not the least afraid of him; but there he beats me’ (p. xxi). Kemble need not have worried. An account of the semi-fraudulent background to much early-nineteenth-century Welsh scholarship, including Guest's work, is given by Bromwich, Rachel, ‘Trioedd Ynys Prydain’ in Welsh Literature and Scholarship (Cardiff, 1969)Google Scholar. History is likely to repeat itself until Anglo-Saxon historians learn to control Celtic evidence for themselves; cf. Dumville, David N., ‘Sub-Roman Britain: History and Legend’, History n.s. 62 (1977), 173–92.CrossRefGoogle Scholar

⁵ Green, John Richard, The Making of England (London, 1881), p. 29Google Scholar. Cf. Burrow, J. W., A Liberal Descent: Victorian Historians and the English Past (Cambridge, 1981), pp. 200–4.CrossRefGoogle Scholar

⁶ Pughe's equation of Andred with an invented Welsh andred ‘uninhabited’, adopted by Guest, has probably helped to form the common notion of the Weald. See Owen[-Pughe], William, A Dictionary of the Welsh Language (London, 1793–1803)Google Scholar, s.vv. andred and tred. Cf. Sawyer, P. H., From Roman Britain to Norman England (London, 1978), pp. 132–3, 136–7 and 139–40.Google Scholar

⁷ The extent of Anglo-Saxon woodland shown on maps has only slowly returned to what is evidenced, despite the devastating criticism of Stevenson, W. H., ‘Dr. Guest and the English Conquest of South Britain’, EHR 17 (1902), 625–42, at 626CrossRefGoogle Scholar; cf. Hill, David, An Atlas of Anglo-Saxon England (Oxford, 1981), pp. 16–17.Google Scholar

⁸ BOH 11, 28: ‘I may refer to, without professing wholly to endorse, the papers of Dr. Guest republished in Origines Celticae, vol. ii, and the early chapters of Mr. Green's Making of England’. Plummer then continues as quoted below, p. 41.

⁹ Stevenson, W. H., ‘The Beginnings of Wessex’, EHR 14 (1899), 32–46, at 46CrossRefGoogle Scholar. Cf. Finnegan, Ruth, ‘A Note on Oral Tradition and Historical Evidence’, History and Theory 9 (1970), 195–201.CrossRefGoogle Scholar

¹⁰ Stenton, F. M., Anglo-Saxon England, 3rd ed. (Oxford, 1971), pp. 21–2Google Scholar; cf. below, pp. 26–7.

¹¹ Kemble, , Saxons in England 1, 3Google Scholar. Kemble seems to have had in mind especially the commonplaces of origin legends (the three ships, the divine ancestor, etc.), but even more important is the tendency for fiction to gather round places and place-names. Stevenson and Stenton seem to have had little feel for this process (cf. below, p. 40), perhaps because the connection between extant Old English poetry (which is mostly set on the continent) and the English landscape is relatively slight (cf. Lapidge, Michael, ‘“Beowulf”, Aldhelm, the “Liber Monstrorum” and Wessex’, SM 3rd ser. 23 (1982), 151–92, at 176–84)Google Scholar; and while English place-name scholars admit the phenomenon of fictitious personal names being created from place-names there is still a tendency to assume that the eponyms of names like Wolferlow and Longslow were real people, not fictitious heroes (cf. e.g. Gelling, Margaret, Signposts to the Past (London, 1978), pp. 154–7 and 188–90)Google Scholar. The much more extensive Celtic literatures supply an illuminating corrective: Flower, Robin, The Irish Tradition (Oxford, 1947), pp. 1–3Google Scholar; Jones, Thomas, ‘The Black Book of Carmarthen “Stanzas of the Graves”’, PBA 53 (1967), 97–137Google Scholar; Padel, O. J., ‘The Cornish Background of the Tristan Stories’, Cambridge Med. Celtic Stud. 1 (Summer 1981), 53–81Google Scholar, esp. 63–4 (with bibliography).

¹² Chadwick, H. Munro, The Origin of the English Nation, 2nd ed. (Cambridge, 1924), pp. 2 and 31.Google Scholar

¹³ Anglo-Saxon England, pp. 717–18.

¹⁴ ‘Gildas and the Anglo-Saxons’.

¹⁵ The History of the Anglo-Saxons, 6th ed. (London, 1836) 1, 158–9.Google Scholar

¹⁶ Although his chief theme was to be English ecclesiastical history, Bede obviously wished to set it in a wider context; why else should he have added his famous passage on the Saxons, Angles and Jutes (see below, p. 24)? This was already apparent to John Richard Green in 1869: see his Historical Studies (London, 1903), p. 338.Google Scholar

¹⁷ Bede would probably have agreed with Kemble on the limitations of oral tradition: ‘All history then, which is founded in any degree upon epical tradition (and national history is usually more or less so founded) must be to that extent imperfect, if not inaccurate; only when corrected by the written references of contemporaneous authors, can we assign any certainty to its records’ (Saxons in England 1, 29). On Bede's opening chapters Kemble writes: ‘Either he could find no more information, or he did not think it worthy of belief. He even speaks doubtfully of the tale of Hengest’ (ibid. p. 27, n. 1).

¹⁸ Sims-Williams, ‘Gildas’, p. 24.

¹⁹ See below, pp. 7, 9 and 11.

²⁰ DEB 26.1. See below, pp. 20 and 38.

²¹ See O'Sullivan, Thomas D., The ‘De Excidio’ of Gildas: its Authenticity and Date, Columbia Stud. in the Classical Tradition 7 (Leiden, 1978)Google Scholar; Sims-Williams, , ‘Gildas’, pp. 2–5Google Scholar; and Gildas: New Approaches, ed. Lapidge, M. and Dumville, D. N. (forthcoming in the series Stud. in Celtic Hist.).Google Scholar

²² DEB 14–20.

²³ See Collingwood, R. G. and Myres, J. N. L., Roman Britain and the English Settlements, 2nd ed. (Oxford, 1937), p. 294Google Scholar; Thompson, E. A., ‘Gildas and the History of Britain’, Britannia 10 (1979), 203–26, at 206–7CrossRefGoogle Scholar; Sims-Williams, , ‘Gildas’, p. 15 and n. 67.Google Scholar

²⁴ Cf. Olrik, Axel, ‘Epic Laws of Folk Narrative’, The Study of Folklore, ed. Alan, Dundes (Englewood Cliffs, N.J., 1965), pp. 129–41, esp. pp. 131, 133–4 and 140Google Scholar. Collingwood's oft-repeated idea that Gildas may reflect a Celtic triad (Roman Britain and the English Settlements, p. 293), though considered possible by Bromwich, Rachel (Trioedd Ynys Prydein, 2nd ed. (Cardiff, 1978), p. lxiii, n. 4, and p. 85)Google Scholar, is not supported by form criticism, for Celtic triads are rarely marked by the sequential and climactic features of the folktale threefold repetitions.

²⁵ DEB 20.1. Winterbottom translates ad Agitium Romanae potestatis virum by ‘to the Roman commander Aëtius’, but this may imply more knowledge than Gildas possessed. Gildas may not even have known that Agitius (if that was the spelling before him) was a spelling of Aëtius.

²⁶ DEB, p. 8; similarly Thurneysen, R., ‘Wann sind die Germanen nach England gekommen?’, Englische Studien 22 (1896), 163–79, at 177Google Scholar. The Latin is quoted below. C. F. C. Hawkes thinks that the Britons were alluding to marine transgressions! See ‘The Jutes of Kent’, Dark-Age Britain: Studies presented to E. T. Leeds, ed. Harden, D. B. (London, 1956), pp. 91–111, at 94Google Scholar. Morris, John elaborates: The Age of Arthur (London, 1973), p. 78.Google Scholar

²⁷ For a reference to Aëtius as ter consul see Martindale, J. R., The Prosopography of the Later Roman Empire 11: A.D. 395–527 (Cambridge, 1980), p. 29.Google Scholar There is no reason to regard ter consuli in DEB 20 as an addition or gloss.

²⁸ The consul of 454 was a different Aëtius (Thompson, ‘Gildas’, p. 215, n. 60; Martindale, , Prosopography 11, 29)Google Scholar.

²⁹ This has been attempted, without prejudice as to the outcome, by D. N. Dumville in chapter 4 of Gildas: New Approaches (see above, n. 21), and the result was as given. I am grateful to Dr Dumville for letting me read a draft.

³⁰ Cf. O'Sullivan, , The ‘De Excidio’, p. 84, n. 52Google Scholar; Hughes, Kathleen, Celtic Britain in the Early Middle Ages, Stud. in Celtic Hist. 2 (Woodbridge, 1980), 91Google Scholar; Sims-Williams, , ‘Gildas’, p. 4.Google Scholar

³¹ Skene, William F., The Four Ancient Books of Wales (Edinburgh, 1868) 1, 35–7Google Scholar; Thurneysen, , ‘Wann sind die Germanen nach England gekommen?’, p. 177.Google Scholar Thurneysen's view was revived by Stevens, C. E. in his influential article ‘Gildas Sapiens’, EHR 56 (1941), 353–73, at 362–3CrossRefGoogle Scholar; cf. O'Sullivan, , The ‘De Excidio’, p. 176Google Scholar. Stevens and the other scholars whose views are summarized by O'Sullivan vie with each other in speculation about the exact date and circumstances of the letter.

³² See the references given by Johnson, Stephen, ‘Late Roman Defences and the Limes’, The Saxon Shore, ed. Johnston, D. E., Council for Brit. Archaeology Research Report 18 (London, 1977), 63–9, at 67.Google Scholar

³³ De Consulatu Stilichonis, 11. 250–5 (Claudii Claudiani Carmina, ed. Birt, Th., Monumenta Germaniae Historica (hereafter MGH), Auct. Antiq. 10 (Berlin, 1892), 211–12)Google Scholar. ‘“Stilicho secured me too”, she said, “when I was being destroyed by neighbouring peoples, when the Scot stirred up all Ireland and the ocean foamed with hostile oars. Such was his care that I did not fear Irish weapons, nor tremble at the Pict, nor look out along my whole shore for the Saxon coming with the uncertain winds.”’

³⁴ DEB 18.3 (presumably Saxons are meant, not Picts and Scots) and 23.2. Cf. Gildas, ed. and trans. Williams, Hugh, Cymmrodorion Record Ser. 3–4 (London, 1899–1901), 52, n. 1Google Scholar; Williams overstates the case for Gildas's knowledge of a Saxon presence in Britain prior to the invitation to the mercenaries, when he suggests that the event of c. 441 discussed below could be subsumed among the raids that had made the Britons fear the Saxons ‘worse than death’.

³⁵ In DEB 14 primum may actually imply that the Irish and Pictish incursions were earlier than the Saxon ones; cf. N. Wright, in Gildas: New Approaches.

³⁶ Some problems are brought out by Hills, Catherine, ‘The Archaeology of Anglo-Saxon England in the Pagan Period: a Review’, ASE 8 (1979), 297–329Google Scholar; Gillam, John, ‘Romano-Saxon Pottery: an Alternative Interpretation’, The End of Roman Britain, ed. Casey, P. J., BAR Brit. ser. 71 (Oxford, 1979), 103–18Google Scholar; Todd, Malcolm, Roman Britain, 55 BC–AD 400, Fontana History of England ([London], 1981), pp. 246–7Google Scholar. The argument that the material ‘collapse of the Roman way of life’ (as seen archaeologically) was not due to the Anglo-Saxons but to the disappearance of demand from the army, etc., leading to economic collapse, is succinctly put by Brown, David, ‘Problems of Continuity’, Anglo-Saxon Settlement and Landscape, ed. Rowley, Trevor, BAR 7 (Oxford, 1974), 16–19.Google Scholar

³⁷ Zosimus, vi. 5.2–3. I quote the translation of Thompson, E. A., ‘Britain, A.D. 406–410’, Britannia 8 (1977), 303–18, at 306CrossRefGoogle Scholar. This is a fundamental account of the chronology of these years; cf. Todd, , Roman Britain, pp. 239–41Google Scholar; Salway, Peter, Roman Britain (Oxford, 1981), pp. 444–5Google Scholar; and Thompson's further discussion in ‘Zosimus 6.10.2 and the Letters of Honorius’, Classical Quarterly n.s. 32 (1982), 445–62.

³⁸ ‘Britanniae Saxonum incursione devastatae’ (Chronica Minora, ed. Mommsen, Th., MGH, Auct. antiq. 9, 11 and 13 (Berlin, 1892–1898) 1, 654)Google Scholar. On the chronology see Thompson, ‘Zosimus 6.10.2’, pp. 453–4. The treatment of the Gallic Chronicle and of Zosimus, vi. 5.3, by Bartholomew, Philip (‘Fifth-Century Facts’, Britannia 13 (1982), 260–70, at 263–4 and 268–70)CrossRefGoogle Scholar is ingenious, but forced in the extreme.

³⁹ Cf. Rivet, A. L. F. and Smith, Colin, The Place-Names of Roman Britain (London, 1979), p. 40Google Scholar; Thomas, Charles, Christianity in Roman Britain to AD 500 (London, 1981), pp. 197 and 200Google Scholar. Professor Thompson points out to me that Britanniae may have become a stereotyped expression for ‘Britain’, as in St Patrick and Constantius.

⁴⁰ Chronica Minora, ed. Mommsen 1, 472.

⁴¹ Vita Prima Sancti Germani, ed. Levison, W., MGH, Script, rer. Meroving. 7 (Hanover and Leipzig, 1920), 225–83Google Scholar (ch. 12). Cf. Williams, Hugh, Christianity in Early Britain (Oxford, 1912), p. 224Google Scholar. The Vita is trans. by Hoare, F. H., The Western Fathers (London, 1954), pp. 281–320.Google Scholar

⁴² Cf. Martindale, , Prosopography 11, 320.Google Scholar

⁴³ Vita Germani, ed. Levison, chs. 17–18. I assume that Germanus and Lupus were not absent from Gaul for more than a year.

⁴⁴ ibid. p. 264, n. The sensible treatment by Williams, (Christianity, pp. 226–7)Google Scholar contrasts favourably with super-subtle readings like that of Myres, J. N. L., Anglo-Saxon Pottery and the Settlement of England (Oxford, 1969), p. 86Google Scholar and n. 2. On the possibility of Germanus having military experience see Martindale, Prosopography 11, 504.

⁴⁵ Note that Constantius omits the Picts in the second sentence quoted above. Despite the loose assumptions of modern writers, Constantius does not give any indication of the distance between the shrine of St Alban (ch. 16) and the battle (or the meeting with the Pelagians). Nor do we know that in 429 the shrine was at Verulamium rather than the place of martyrdom, which seems to have been near the Thames (see DEB 11.1; Sims-Williams, ‘Gildas’, p. 6, n. 25).

⁴⁶ Vita Germani, ed. Levison, chs. 25–7. The most recently published study of the chronology of Germanus’ life is Ralph Mathisen, W., ‘The Last Year of Saint Germanus of Auxerre’, AB 99 (1981), 151–9Google Scholar; but new studies are forthcoming from Ian Wood (in Gildas: New Approaches, ed. Lapidge and Dumville) and E. A. Thompson. The latter kindly tells me that he places the second mission in 437.

⁴⁷ ‘Brittanniae usque ad hoc tempus variis cladibus eventibusque latae [laceratae?] in dicionem Saxonum rediguntur’ (Chronica Minora, ed. Mommsen 1, 660). Cf. the ‘Chronicler of 511’: ‘Britanniae a Romanis amissae in dicionem Saxonum cedunt’ (ibid. p. 661). The chronology offers problems (cf. Thompson, ‘Zosimus 6.10.2’, pp. 453–4), but it seems safest to rely on the regnal dates (the system common to both chronicles). The ‘Chronicler of 452’ gives his entry in the nineteenth year of the twenty-seven years of Theodosius (apparently 424–50) – or, according to Miller, M. (‘The Last British Entry in the “Gallic Chronicles”’, Britannia 9 (1978), 315–18, at 317)CrossRefGoogle Scholar, spread over the eighteenth to nineteenth years – while the ‘Chronicler of 511’ gives his in the sixteenth year of the twenty-five years of Theodosius and Valentinian (apparently 426–50), i.e. they apparently agree on c. 441–2. Miller's notion that 445–6 is meant in the ‘Chronicle of 452’, and that this date was derived from Bede, along with the date of Magnus Maximus’ usurpation (cf. HE v.24), depends on fantastic premises. Cf. Casey, P. J., ‘Magnus Maximus in Britain’, The End of Roman Britain, ed. Casey, , pp. 66–79, at 70–2 and 78, n. 17Google Scholar. Miller's views (e.g., on Britanniae being an anachronism, but see above, p. 10, n. 39) are elaborated by Bartholomew, ‘Fifth-Century Facts’, pp. 268–70.

⁴⁸ For various interpretations see Harrison, , Framework, pp. 26–9.Google Scholar With the wording cf. Ammianus Marcellinus, 28.3.7, on the province of Valentia in 369, ‘quae in dicionem concesserat hostium, ita reddiderat [Theodosius] statui pristino’ (ed. J. C. Rolfe (London, 1935–9) 111, 134). Alfred Anscombe proposed that ‘What we have here is the record of a diplomatic arrangement; the Britannias were placed under the authority of the Saxons, no doubt by Aëtius', (‘The Richborough Coin Inscribed “Domino Censaurio Ces”’, BNJ 19 (1927–1928), 1–23, at 15, n. 1)Google Scholar. This is a guess, but would fit in with Aëtius’ policy on the continent, as described by the Chronicler himself (cf. Salway, , Roman Britain, p. 473)Google Scholar; yet why should his wording be so vague? (Compare the following entry on Aëtius' Alan settlement in Brittany.) And is it not strange that Gildas has no memory of such a sell-out (the blunder of his superbus tyrannus being of a different order)? Yet if Anscombe is right, the ‘Saxon’ interpretation of the letter to Aëtius seems even more likely; cf. the Armoricans' protests about the Alan settlement (Vita Germani, ed. Levison, ch. 28).

⁴⁹ The story of the Easter victory of Germanus in 429 or 430 only becomes important in the light of the events of c. 408 and c. 441; otherwise one might treat it on the same level as the incursions of barbarians (which included Saxons) of the 360s. On these see Ammianus Marcellinus, 26.4.5 and 27.8.1–10, ed. and trans. Rolfe 11, 588–9, and 111, 50–7; cf. Casey, ‘Magnus Maximus’, pp. 73–4 and 78, n. 20a.

⁵⁰ ‘Gildas’, p. 215.

⁵¹ ibid. p. 220; cf. p. 216, n. 63, and Harrison, , Framework, pp. 23–4.Google Scholar

⁵² Gildas's aim was to generalize about all Britannia, though of course he has to dwell on certain regions at times; cf. Sims-Williams, ‘Gildas’, p. 7.

⁵³ DEB 20.1.

⁵⁴ Cf. Meid, Wolfgang, Die Romanze von Froech und Findabair, Innsbrucker Beiträge zur Kulturwissenschaft, Sonderheft 30 (Innsbruck, 1970), 202Google Scholar; Hamp, Eric P., ‘On the Fundamental IE Orientation’, Ériu 25 (1974), 253–61.Google Scholar

⁵⁵ The possibility of this emendation was pointed out to me by Dr Lapidge.

⁵⁶ DEB 4.4; I use the translation of Williams, , Gildas, p. 19.Google Scholar On this passage cf. Green, , Historical Studies, p. 6Google Scholar; Thompson, ‘Gildas’, pp. 209 and 223, and his review of Winterbottom's, edition, Britannia 11 (1980), 452Google Scholar; and Sims-Williams, , ‘Gildas’, p. 7.Google Scholar

⁵⁷ They could not be dated because the northern troubles were perennial. Bede placed them after the sack of Rome in 410 (see below), but Gildas evinces no knowledge of that event, nor of any ‘Honorian Rescript’ (cf. Sims-Williams, ‘Gildas’, p. 17).

⁵⁸ Cf. Blair, Peter Hunter, The Origins of Northumbria (repr. from A Ae 4th ser. 25) (Gateshead on Tyne, 1948), p. 14Google Scholar. ‘Aetio III et Symmaco’ are correctly given as consuls sub anno Passionis 419 in the Victorine Easter Tables (ed. Mommsen, Chronica Minora 1, 722; re-ed. Bruno Krusch, Studien zur christlich-mittelalterlichen Chronologie, Abhandlungen der Preus-sischen Akademie der Wissenschaften 1937, Phil.-hist. Kl. 8 (Berlin, 1938), 47). These tables may not have begun to attract interest in the Insular churches until after the Council of Orléans in 541, when they were officially adopted in Gaul; cf. Harrison, , Framework, p. 33Google Scholar; idem, ‘Episodes in the History of Easter Cycles in Ireland’, Ireland in Early Mediaeval Europe: Studies in memory of Kathleen Hughes, ed. Whitelock, Dorothy et al. (Cambridge, 1982), pp. 307–19, esp. pp. 307, 309–10 and 318Google Scholar. Columbanus deplores them in his letter to Pope Gregory c. 600 (ibid. p. 310, n. 16), but though he speaks favourably of Gildas in another context in the letter, there is no tradition here or anywhere else that Gildas took part in the Easter controversy; he may have lived too early. Johnstone's, P. K. suggestion of ‘A Consular Chronology of Dark Age Britain’ (Antiquity 36 (1962), 102–9)CrossRefGoogle Scholar depends on an unlikely view of the Historia Brittonum (cf. Dumville, David N., ‘Some Aspects of the Chronology of the Historia Brittonum’, BBCS 25 (1972–1974), 439–45Google Scholar) and a single inscription from Penmachno, about which doubts have recently been raised by Mr Richard White of the Gwynedd Archaeological Trust.

⁵⁹ DEB 1.2, 23.3 and 26.1.

⁶⁰ Both works are ed. Mommsen, T., Chronica Minora 111, 223–354Google Scholar. Their reckoning is by Anno Mundi, the birth of Christ being A.M. 3952 (ibid. pp. 281–2); cf. Jones, Charles W., Saints' Lives and Chronicles in Early England (Ithaca, N.Y., 1947), p. 26Google Scholar; and Miller, M., ‘Bede's Roman Dates’, Classica et Mediaevalia 31 (1970), 239–52Google Scholar. In this and the next paragraph I give true a.d. dates (in brackets) for the sake of clarity; I do not attempt to answer the merely hypothetical question ‘what a.d. dates would Bede have set down had he been using a.d. dating in the Chronica Maiora?’.

⁶¹ Chronica Maiora, §§457–61.

⁶² ibid. §§469 and 473–4; HE 1.11–12 and v.24. He dates the sack of Rome a year too early (a.d. 409) in HE. Jones's statement that in the Chronica Maiora Bede ‘had placed it after a.m. 4377, which would make the event occur after a.d. 425’ (Saints' Lives, p. 33), results from misunderstanding the placing of the a.m. obits, which stand at the beginnings of the reigns of the emperors to whom they refer!

⁶³ Chronica Maiora, §§483–4. The form Vertigernus is more archaic than the forms in Vor- and Gur- found in some texts of Gildas and the form in Uur- in HE 1.14; see Jackson, K. apud Dumville, David N., ‘A New Chronicle-Fragment of Early British History’, EHR 88 (1973), 312–14, at 313, n. 5Google Scholar. On the probability that Gildas meant Vertigern/Vortigern, even if he did not name him, see Jackson, Kenneth, ‘Varia, 11: Gildas and the Names of the British Princes’, Cambridge Med. Celtic Stud. 3 (Summer 1982), 30–40, at 35–40.Google Scholar

⁶⁴ It has been suggested that the fact that Bede spreads Gildas's account of the invitation to the mercenaries (glossed Angli by Bede) and the arrival of the ‘gens Anglorum sive Saxonum … tribus longis navibus’ over the reigns of Theodosius and Marcian (§§484 and 489) reflects a ‘conception of a composite Adventus’ (Myres, J. N. L., ‘The Adventus Saxonum’, Aspects of Archaeology in Britain and Beyond: Essays presented to O. G. S. Crawford, ed. Grimes, W. F. (London, 1951), pp. 221–41, at 233)Google Scholar. This is wrong; Bede never questions that the first adventus was the one in three ships.

⁶⁵ For example, it could already have been inserted in the imperial list that he used as his framework. But this is only a possibility. It is often wrongly argued that the Marcian synchronism was pre-Bedan (and perhaps Kentish, though there is no sign of the addition of Kentish material in the Chronica Maiora where it first appears) because it conflicts with Bede's own (alleged) 446 or 446/7 date for the adventus: Chadwick, , Origin, p. 46Google Scholar; Myres, ‘The Adventus Saxonum’, pp. 225 and 230; Hawkes, ‘The Jutes of Kent’, p. 92; Harrison, , Framework, pp. 125 and 140Google Scholar. But Bede's alleged belief in 446 or 446/7 is an illusion (see ibid. p. 19; BOH 11, 27, and below) so there is no conflict. The suggestion of Dumville (‘Some Aspects’, p. 444, n. 3) is open to the objection voiced by Chadwick, , Origin, p. 46Google Scholar, n. 2 (cf. Guest, , Origines Celticae 11, 169).Google Scholar

⁶⁶ Chronica Maiora, §491 (= Constantius, Vita Germani, ed. Levison, chs. 12–18) and §492 (= Constantius, chs. 28–46). These chapters are sandwiched between the Invention of the Baptist's head (a.d. 453) and Valentinian's murder of Aëtius (a.d. 454), both taken from the Chronicle of Marcellinus (ed. Mommsen, Chronica Minora 11, 84–6); Bede would have been able to verify the dates of these two events by comparing Marcellinus' consuls with the Victorine Easter Tables for a.p. 426–7 (for editions see above, p. 14, n. 58).

⁶⁷ Cf. Chronica Maiora, §§478–9 (based on Marcellinus’ entries for the equivalent of a.d. 424–5). Bede would not have known that Placidia died in 450.

⁶⁸ Vita Lupi Episcopi Trecensis, ed. Krusch, B., MGH, Script. rer. Meroving. 7 (Hanover and Leipzig, 1920), 284–302Google Scholar (chs. 4–5) (my italics). Cf. Prosper, , s.a. Passionis 424 (ed. Mommsen, , Chronica Minora 1, 481–2)Google Scholar. As Krusch notes (pp. 284–5, 288 and 302, n. 1), Bede quotes the Vita Lupi in HE 1.21.

⁶⁹ Thus HE 1.21. Bede's sources, Marcellinus and Prosper (ed. Mommsen, Chronica Minora 1, 483–4, and 11, 86), correctly place the murder in a consular year equivalent to 455. Bede could have confirmed this by Marcellinus' indiction and Prosper's a.p. date.

⁷⁰ Chronica Maiora, §541: ‘adventus autem Anglorum in Brittaniam plus minus anno CLXXX’. Cf. HE 11.14 and v.24 s.a. 627.

⁷¹ HE v. 23. Cf. 1.23, 11. 14; and ‘The Moore Memoranda on Northumbrian History’, ed. Blair, Peter Hunter, The Early Cultures of North-West Europe (H. M. Chadwick Memorial Studies), ed. Fox, Cyril and Dickins, Bruce (Cambridge, 1950), pp. 245–57, at 246 and 256Google Scholar. There is no conflict between Bede's imperial bracket and his approximations, as often claimed (cf. above, p. 16, n. 65).

⁷² Whitelock, , Documents, p. 4, n. 7.Google Scholar

⁷³ Chronica Maiora, §504. Cf. Miller, M., ‘Bede's Use of Gildas’, EHR 90 (1975), 241–61CrossRefGoogle Scholar. For the sake of brevity I do not register points of agreement or disagreement with this ambitious attempt to combine history and historiography.

⁷⁴ HE 1.5 and 12. See the notes by Plummer and by Colgrave and Mynors in their editions; and Blair, Hunter, Origins of Northumbria, pp. 15–16.Google Scholar

⁷⁵ HE 1.17–21. This ‘flashback’ begins ‘a few years’ before the adventus with the Pelagian heresy (from Prosper, but omitting his date) but ends with the sixth year of Marcian. Bede may have guessed either that Germanus and Lupus arrived some years after the outbreak of the heresy or that they stayed some years in Britain (or both); either way they could have been there after the adventus to meet the Saxons and Picts. (Bede does not say that Germanus' victory was before the adventus.) In Chronica Maiora, §489, Bede has Gildas's Saxon reinforcements (DEB 23.4) drive off the Irish and Picts, then attack the Britons (‘primo hostes quos petebatur abigit: deinde in socios arma vertens…’); in HE 1.15 he expands this: ‘Turn subito inito ad tempus foedere cum Pictis, quos longius iam bellando pepulerant, in socios arma uertere incipiunt’ (cf. the chapter heading: ‘Ut inuitata Britanniam gens Anglorum primo quidem aduersarios longius eiecerit, sed non multo post iuncto cum his foedere in socios arma uerterit’). This is surely a deduction from and allusion to the ‘bellum Saxonum Pictorumque adversus Brittones eo tempore iunctis viribus susceptum’ which Bede took from Constantius (Chronica Maiora, §491; cf. HE 1.20). I disagree here with Blair, Hunter, Origins of Northumbria, pp. 18–19.Google Scholar

⁷⁶ HE 1.13. Bede dates the appeal to the year after Attila's fratricide, which his source, Marcellinus (ed. Mommsen, Chronica Minora 11, 81), places in the consular year preceding Aëtius' third consulship. (Prosper places it two years before it: ibid. 1, 480.) Bede must have known that Aëtius' third consulship was in 446 from Marcellinus' indiction, from Prosper's Annus Passionis, and from the Victorine Easter Tables (despite Miller, ‘Bede's Roman Dates’, p. 250); but he does not give an a.d. date for it.

⁷⁷ Chronica Minora, ed. Mommsen, 11, 80–2Google Scholar; HE 1.13–14; DEB 2, 20.2 and 22.2 (Bede clearly did not take interea in DEB 20.2 in the forced sense advocated by Miller, ‘Bede's Use of Gildas’, p. 260.) Gildas rather implies that the British famine was due to Irish and Pictish raids (DEB 19.4). Bede distinguishes the British famine by the phrase fames sua praefata (HE 1.14); the omission of sua in the ‘c2’ group of manuscripts suggests that already in the eighth century someone thought he knew better than Bede.

⁷⁸ Cf. the criticisms of Stevens's attempt to identify Gildas's pestifera lues (‘Gildas Sapiens’, p. 363) made by Myres (‘The Adventus Saxonum’, pp. 228–9), Harrison, (Framework, pp. 25–6)Google Scholar and Todd, Malcolm (‘Famosa Pestis and Britain in the Fifth Century’, Britannia 8 (1977), 319–25)CrossRefGoogle ScholarPubMed. Todd's remark (p. 320) that Gildas implies a date before c. 445 is a slip.

⁷⁹ HE 1.15 and v.24; Martindale, , Prosopography 11, 714–15Google Scholar. Bede's 449 is difficult to explain (Harrison, , Framework, p. 126, n. 7)Google Scholar. Jones, (Saints' Lives, p. 34Google Scholar) links it with Bede's antedating of the sack of Rome and the Palladian mission (HE v.24, s.aa. 409 and 430).

⁸⁰ See above, p. 18, n. 69.

⁸¹ The wording in HE 1.15 does not require the adventus to be before the death of Valentinian.

⁸² DEB 26.1. See the survey of scholarship in O'Sullivan, , The ‘De Excidio’, pp. 134–57Google Scholar (to which add: Winterbottom, Michael, ‘Notes on the Text of Gildas’, JTS n.s. 27 (1976), 132–40, at 135).CrossRefGoogle Scholar

⁸³ BOH 11, 31. Coincidences do after all occur – what is the probability of the present writer having been born 1500 years after 449 (as is the case)? – yet O'Sullivan describes Plummer's view as ‘eccentric’ and ‘far-fetched’ (The ‘De Excidio’, pp. 139 and 156).

⁸⁴ Chronica Maiora, §504.

⁸⁵ This conventional approximation is ultimately based either on the 446 Aëtius synchronism, or on the Welsh obit of Maglocunus (on which see above, p. 6), or on both, or, of course, on Bede.

⁸⁶ There is indeed a coincidence, and one that Bede may have relished: that 500 minus 456 equals 44, the very figure given by Gildas in the context of Badonicus mons but calculating forwards in time! (Strictly, the forty-fourth year is 499, counting inclusively.)

⁸⁷ HE 1.15. This sentence follows a probable authorial interpolation on Angles, Saxons and Jutes, as Myres showed (see below, p. 24). This probability undermines Myres's further idea that Bede meant that Hengest and Horsa were the leaders of the (supposedly Jutish, etc.) reinforcements (‘The Adventus Saxonum’, pp. 232–3; Hawkes, ‘Jutes of Kent’, p. 93); this is mere speculation (see Harrison, , Framework, p. 21Google Scholar, n. 9) and is contradicted by HE 11.5. In any case Gildas's obvious implication is that both the first arrivals and their reinforcements arrived and were settled in eastern Britain, not south-east England where they would have been manifestly useless against the northern enemies. ‘Vortigern's need for protection against invasion from Roman Gaul’ (Myres, , Pottery and the Settlement, p. 96Google Scholar) is romance, not inference.

⁸⁸ BOH 1, xlv, and 11, 4 and 28; cf. above, p. 5, n. 17.

⁸⁹ Collingwood, and Myres, , Roman Britain and the English Settlements, p. 358, n. 1Google Scholar; Morris, , Age of Arthur, p. 578Google Scholar. The form Hors occurs in the Historia Brittonum; cf. Chadwick, , Origin, p. 43 and n. 1Google Scholar. (Is the lapis tituli ‘qui est super ripam Gallici maris’ of ch. 43 (ed. Mommsen, Chronica Minora 111, 187–8) the Bedan monument?) With Bede's words (‘Horsa postea occisus est in bello a Brettonibus hactenus in orientalibus Cantiae partibus monumentum habet suo nomine insigne’) cf. HE 111.4 (‘sedem episcopatus, sancti Martini episcopi nomine et ecclesia insignem’).

⁹⁰ HE 11.5. Cf. above, p. 16, n. 63. The form Uurtigernus doubtless came via Kent.

⁹¹ If so, Oeric may have been so named to alliterate with the ancestral Oisc.

⁹² On early study of DEB see Dumville, ‘Sub-Roman Britain’, pp. 183–4.

⁹³ ‘In oceano vero occidentale est insula quae dicitur Britania, ubi olim gens Saxonum veniens ab antiqua Saxonia cum principe suo nomine Ansehis [var. Ansehys] modo habitare videtur’ (ed. Schnetz, Joseph, Itineraria Romana 11 (Leipzig, 1940), 105Google Scholar. Cf. the plates (not text) of Richmond, I. A. and Crawford, O. G. S., ‘The British Section of the Ravenna Cosmography’, Archaeologia 93 (1949), 1–50.CrossRefGoogle Scholar) On the date and possibility of interpolation see Rivet, and Smith, , Place-Names of Roman Britain, p. 185Google Scholar, n. 1, and my review in Cambridge Med. Celtic Stud. 4 (Winter 1982), 90–3, at 91.Google Scholar

⁹⁴ ibid. p. 91, following Chadwick, , Origin, p. 45Google Scholar; for other discussions see references in Turville-Petre, J. E., ‘Hengest and Horsa’, SBVS 14 (1953–1957), 273–90, at 285, n. 42Google Scholar. In the Cosmography, ch for non-assibilated [k] is common (Schnetz, Joseph, ‘Untersuchungen über die Quellen der Kosmographie des anonymen Geographen von Ravenna’, Sitzungsberichte der bayerischen Akademie der Wissenschaften, Phil.-hist. Kl. 6 (Munich, 1942), 36)Google Scholar. For the sound changes that would give Oisc <* Anskiz see Campbell, A., Old English Grammar (Oxford, 1959)Google Scholar, §§121, 198 and 404. Campbell is wisely silent on the chronology, but even if *Anschis is a transcription of something like [ ski] it is unlikely that such a form would be heard from English lips far into the sixth century. Yet, if the story of Oisc was current on the other side of the Channel, it might have been picked up later on from an Old Saxon speaker with [s-], or from another Germanic dialect with [ans-]. Another possibility is that Oisc or similar was ‘translated’ Ansc(h)is, by sound-substitution or analogy, by a continental intermediary; for this phenomenon cf. BOH 11, 291; Jacobs, Nicolas, ‘Anglo-Danish Relations, Poetic Archaism and the Date of Beowulf’, Poetica (Tokyo) 8 (Autumn 1978), 23–43, at 28Google Scholar; Sims-Williams, Patrick, ‘An Unpublished Seventh- or Eighth-Century Anglo-Latin Letter in Boulogne-sur-Mer MS 74 (82)’, MÆ 48 (1979), 1–22, at 7.Google Scholar

⁹⁵ Getica, §78, ed. Mommsen, Th., MGH, Auct. antiq. 5.i (Berlin, 1882), 76Google Scholar. Cf. Turville-Petre, ‘Hengest and Horsa’, p. 284; Moisl, Hermann, ‘Anglo-Saxon Royal Genealogies and Germanic Oral Tradition’, JMH 7 (1981), 215–48, at 219–23 and 235–6Google Scholar; Faulkes, Anthony, ‘Descent from the Gods’, MScand 11 (1978–1979), 92–125Google Scholar. The Cosmographer may reflect earlier interest in the Mediterranean world in origines gentium akin to that satisfied by Jordanes. The Kentish place-name Oesuualun (Easole) is of interest in relation to the above. See Smith, A. H., English Place-Name Elements, EPNS 25–6 (Cambridge, 1956) 1, 159–60.Google Scholar

⁹⁶ Turville-Petre, ‘Hengest and Horsa’; de Vries, Jan, ‘Die Ursprungssage der Sachsen’, Niedersächsisches Jahrbuch für Landesgeschichte 31 (1959), 20–37, at 30–2Google Scholar. Cf. Olrik, ‘Epic Laws of Folk Narrative’, p. 140. The brothers Eofor and Wulf in Beowulf, cited by Chadwick, (Origin, p. 43Google Scholar, n. 1), are fictional of course.

⁹⁷ ibid. p. 49; Sisam, Kenneth, ‘Anglo-Saxon Royal Genealogies’, PBA 39 (1953), 287–348, at 326Google Scholar; Chambers, R. W., ‘Beowulf’: an Introduction, 3rd ed. (Cambridge, 1959), pp. 443–4, 538 and 544–5.Google Scholar

⁹⁸ Felix's Life of Saint Guthlac, ed. and trans. Colgrave, Bertram (Cambridge, 1956)Google Scholar, chs. 1–2; Sisam, ‘Anglo-Saxon Royal Genealogies’, pp. 306–7 and 329. A wonderfully imaginative account of Hengest and Icel is given by Morris, , Age of Arthur, pp. 266–73.Google Scholar

⁹⁹ HE 1.15 (‘Aduenerant…orti’). See Collingwood, and Myres, , Roman Britain and the English Settlements, p. 337Google Scholar, n. 1 (not all Myres's arguments are convincing; cf. above, p. 21, n. 87). For the probability of a written source, see Campbell, James, Bede's ‘Reges’ and ‘Principes’, Jarrow Lecture 1979 [Jarrow, 1980], p. 13, n. 9.Google Scholar

¹⁰⁰ Wars viii. 20. For translations see Thompson, E. A., ‘Procopius on Brittia and Britannia’, Classical Quarterly, n.s. 30 (1980), 498–507CrossRefGoogle Scholar; and Rivet, and Smith, , Place-Names of Roman Britain, p. 83Google Scholar. With the idea of one king per nation cf. vi.15 on Thule (ibid. p. 82). Brittia is Britain, but what Procopius says about Frankish claims to it may really relate to Brittany (see Thompson, pp. 501–2; note that Breton Breiz ‘Brittany’ is < * Brittiā (Jackson, Kenneth Hurlstone, A Historical Phonology of Breton (Dublin, 1967), p. 93)Google Scholar, whereas Welsh Prydain ‘Britain’ is <* Pritaniā). Is Brettania, which Procopius distinguishes from the island of Brittia, Ptolemy's Mikra Brettania, that is, Ireland (cf. Rivet and Smith, p. 112)? J. N. L. Myres's view that Bede's list of Germanic tribes, including Frisians, in HE v.9, refers to non-Anglo-Saxon elements in the English population (‘The Angles, the Saxons, and the Jutes’, PBA 56 (1970), 145–74, at 151, n. 2Google Scholar) is due to misreading. On Bede's Garmani see Hamp, Eric P., ‘Latin er in British Celtic’, Études celtiques 17 (1980), 161–3.Google Scholar

¹⁰¹ Colin Smith points out that Winta son of Woden in the royal genealogy of Lindsey (not East Anglia) may be an eponym of Venta Icenorum (Caistor St Edmund, Norfolk). See his ‘The Survival of Romano-British Toponymy’, Nomina 4 (1980), 27–40, at 36Google Scholar (on p. 40 he compares Dimet (Dyfed) son of Magnus Maximus in the Welsh Demetian pedigree); cf. Preparatory to ‘Anglo-Saxon England’: Being the Collected Papers of Frank Merry Stenton, ed. Stenton, Doris Mary (Oxford, 1970), p. 128.Google Scholar

¹⁰² The claim of the Hampshire ‘Jutes’ to the name is obviously better supported than the Kentish claim (cf. HE iv. 16; Chadwick, , Origin, pp. 3–5, 26–8, 84 and 97)Google Scholar. If the Ravenna Cosmographer was correctly informed, the Oiscingas of Kent claimed Saxon descent.

¹⁰³ Bede uses virtually none of Gildas's material on the British church (see HE 1.22). In his view it was irrelevant to English church-history.

¹⁰⁴ HE v.24. Cf. BOH 11, 141; Two of the Saxon Chronicles Parallel, ed. Plummer, Charles (Oxford, 1892–1899) 11, 14Google Scholar. Is it suspicious that 547 is a round figure in A.P. dating (DXX) and that the reigns of Ida and Æthelfrith (HE 1.34) lasted for the conventional-looking figures of 12 and 24?

¹⁰⁵ HE 11.5. See the admirably sane comments on this much misused passage by Kemble, , Saxons in England 11, 8–22Google Scholar; also Bately, Janet, ‘Bede and the Anglo-Saxon Chronicle’, Saints, Scholars and Heroes: Studies in Medieval Culture in Honour of Charles W. Jones, ed. King, Margot H. and Stevens, Wesley M. (Collegeville, Minn., 1979) 1, 233–54, at 243)Google Scholar; and Alfred the Great: Asser's Life of Alfred, and other Contemporary Sources, trans. Keynes, Simon and Lapidge, Michael (Harmondsworth, 1983), p. 210, n. 2.Google Scholar

¹⁰⁶ An interesting recent discussion is perhaps misleading on this point: Campbell, James, John, Eric and Wormald, Patrick, The Anglo-Saxons (Oxford, 1982), p. 53Google Scholar (see also pp. 54, 73, 99–100 and 249).

¹⁰⁷ Sweet, Henry, ‘Some of the Sources of the Anglo-Saxon Chronicle’, Englische Studien 2 (1879), 310–12Google Scholar; Sisam, K. apud Aileen, and Fox, Cyril, ‘Wansdyke Reconsidered’, ArchJ 115 (1960), 1–48, at 46Google Scholar; cf. Bately, Janet, ‘The Compilation of the Anglo-Saxon Chronicle, 60 bc to ad 890: Vocabulary as Evidence’, PBA 64 (1978), 93–129, at 102–3Google Scholar. The ‘poetic’ imagery of the 473 annal may reflect Gildas rather than an Old English source.

¹⁰⁸ Kemble, , Saxons in England 1, 32–3.Google Scholar

¹⁰⁹ Bately, ‘Compilation’.

¹¹⁰ Chadwick, , Origin, p. 25Google Scholar; Stenton, , Preparatory to ‘Anglo-Saxon England’, pp. 119–20Google Scholar; Harrison, , Framework, pp. 133–4 and 140Google Scholar; Dumville, David, ‘Latin and Irish in the Annals of Ulster, a.d. 431–1050’, Ireland in Early Mediaeval Europe, ed. Whitelock et al., pp. 320–41, at 333–4.Google Scholar

¹¹¹ Stenton blatantly speculates: Preparatory to ‘Anglo-Saxon England’, pp. 120–1. Some dubious philological evidence is given by W. H. Stevenson, ‘Beginnings of Wessex’, pp. 37–8 (cf. Bately, ‘Compilation’, p. 105, n. 4). On Wihtgara see Chadwick, , Origin, p. 21, n. 1Google Scholar; and Campbell, , Grammar, p. 247Google Scholar. The spelling of Giwis(ing) may reflect Welsh influence rather than archaism; see Sisam, ‘Anglo-Saxon Royal Genealogies’, p. 303, n. 2; and Armes Prydein, ed. and trans. Williams, Ifor and Bromwich, Rachel (Dublin, 1972), pp. xv–xvi and 49–50Google Scholar. Finally, on the semantics of lēag see Gelling, Margaret, ‘Some Notes on Warwickshire Place-Names’, Birmingham and Warwickshire Archaeol. Soc. Trans. 86 (1974), 59–79, at 67–8Google Scholar; and Johansson, Christer, Old English Place-Names and Field-Names containing lēah, Stockholm Stud. in Eng. 32 (Stockholm, 1975)Google Scholar. Stenton remarks that some of the Kentish names were unfamiliar to the West Saxon Chronicler (Anglo-Saxon England, pp. 16–17); this does not make them necessarily archaic.

¹¹² On the possibility of Kentish records in Æthelberht's time see Harrison, , Framework, pp. 121–3Google Scholar. On the 538 and 540 entries see Jones, , Saints' Lives, p. 34.Google Scholar

¹¹³ Campbell, J., ‘Bede's Words for Places’, Names, Words, and Graves, ed. Sawyer, P. H. (Leeds, 1979). pp. 34–54, at 50, n. 20.Google Scholar

¹¹⁴ Leslie Alcock shrewdly points out that the name ‘Dyrham Camp’ is modern, the older name of the hillfort being Burrill (1695, <*burh-hyll), so that the Chronicler is not necessarily referring to a hillfort site (‘Her…gefeaht wip Walas: Aspects of the Warfare of Saxons and Britons’, BBCS 27 (1976–1978), 413–24, at 421–2Google Scholar; cf. Royal Commission on Historical Monuments, County of Gloucester 1: Iron Age and Romano-British Monuments in the Gloucestershire Cotswolds (London, 1976), p. 51)Google Scholar; on the other hand if the burh had no distinctive name, the Chronicler might be constrained to call it Deorham. The fact that battles have been fought at hillforts, river-crossings, etc., throughout English history (down to the Civil War) provides no support for the Chronicle; it offers a background against which stories about battles at such sites may have arisen and have been attached to more or less legendary heroic names; for example, traditions of the 715 battle at ‘Woden's Barrow’ may have influenced those of the ‘592’ battle, and so on.

¹¹⁵ The Archaeology of the Anglo-Saxon Settlements, repr. with intro. by Myres, J. N. L. (Oxford, 1970)Google Scholar.

¹¹⁶ Stevens, ‘Gildas Sapiens’, p. 369; Evison, Vera I., The Fifth-Century Invasions South of the Thames (London, 1965), p. 84Google Scholar. Cf. Harrison, , Framework, pp. 130–1Google Scholar; and Alcock, , ‘Her…gefeaht’, p. 414.Google Scholar

¹¹⁷ Leeds, , Archaeology, p. 52.Google Scholar

¹¹⁸ Myres, J. N. L., ‘Wansdyke and the Origin of Wessex’, Essays in British History presented to Sir Keith Feiling, ed. Trevor-Roper, H. R. (London, 1964), pp. 1–27Google Scholar (cf. Collingwood, and Myres, , Roman Britain and the English Settlements, p. 404, n. 3)Google Scholar; Kirby, D. P., ‘Problems of Early West Saxon History’, EHR 80 (1965), 10–29CrossRefGoogle Scholar, esp. 22–4 and 28. I do not dispute that the ninth-century and later sources for the sixth-century dynasts are confused (indeed, hopelessly confused); I dispute only that we can profit from their confusion to divine the true state of affairs: the sources do not portray Ceawlin as anything but a Cerdicing, and I do not see that we can penetrate beyond them to any earlier belief.

¹¹⁹ In Myres's case (‘Wansdyke’, p. 23) a mere scribal error in the Parker Chronicle (see Whitelock, , Documents, p. 146, n. 3)Google Scholar; in Kirby's case (‘Problems’, p. 24) the fact that Ceawlin's genealogy ‘is never given in the annalistic entries relating to him, and the first pedigree to name Cynric as the father of Ceawlin is that of Cædwalla (sub 685), which shows some sign of being a seventh-century genealogy’ (would that one could date it so early!).

¹²⁰ Despite Myres (‘Wansdyke’, pp. 23–4) these contain not Cūtha but Cūthen, which could stand for any of the Cūth- names which occur in all branches of the royal house of Wessex, or for none of them, seeing that the name-element occurs throughout England. See Ekwall, Eilert, The Concise Oxford Dictionary of English Place-Names, 4th ed. (Oxford, 1960), pp. 135 and 137Google Scholar; Dickinson, Tania M., Cuddesdon and Dorchester-on-Thames, BAR 1 (Oxford, 1974), 32–4Google Scholar; cf. Morris, , Age of Arthur, pp. 294–5 and 630.Google Scholar

¹²¹ For further speculations see Fox and Fox, ‘Wansdyke Reconsidered’; Green, H. Stephen, ‘Wansdyke, Excavations 1966 to 1970’, Wiltshire Archaeol. and Nat. Hist. Mag. 66 (1971), 129–46Google Scholar. All that is certain about East Wansdyke is that it is late or post-Roman. The notion that the name Wodnesdic shows anything (e.g. that it is of the pagan Saxon period) is false; the Woden tag may have been transferred to the Dyke from Wodnesbeorg at a comparatively late date.

¹²² Cf. Plummer, , Two Chronicles 11, 10Google Scholar: ‘If the Saxons really landed there, then the origin alike of our nationality and of our Christianity is closely bound up with that little spot.’

¹²³ On these see Chadwick, , Origin, pp. 36–50.Google Scholar

¹²⁴ Cf. above, p. 21 and n. 89. The usual identification with Aylesford (on the east of the Medway, which suits Bede's wording) is supported by ‘Episford’ in the Historia Brittonum, ch. 44 (ed. Mommsen, Chronica Minora 111, 187; the first element of the name may have been confused in the latter with the British stem ep- ‘horse’, seen in Welsh ebol, epil). The most striking monument here is Kit's Coty House; this or one of the neighbouring megalithic monuments could have been associated with Horsa through a folk-etymology of Horsted (on the Roman road to Rochester, whence Bede's story may have come). They are described in the Victoria County History of Kent 1 (London, 1908), 318–20Google Scholar. Lambard(e), William, in A Perambulation of Kent (London, 1576)Google Scholar, surmised that Horsted is named from Horsa (pp. 288–9; fuller on pp. 363–4 and 409–10 of the 1596 ed.), and in this was supported by Camden, William on p. 169 of his Britannia (London, 1586)Google Scholar. John Stow visited the area in 1590, and in an addition to the 1592 ed. of The Annales of England claimed that the ‘people of that countrey say the sayd Horse was slaine’ at Horsted (p. 55). By 1763 a heap of flints at Horsted was superstitiously regarded as the remains of Horsa's monument ‘by the country people’ (MrColebrooke, , ‘An Account of the Monument Commonly Ascribed to Catigern’, Archaeologia 2 (1773), 107–17, at 110–11)CrossRefGoogle Scholar. The folk belief may well be of learned inspiration, yet I suspect that it may give us a better idea of the true character of the Chronicle entry than any amount of speculation on Hengest's strategy.

¹²⁵ Gelling, Margaret, ‘Latin Loan-Words in Old English Place-Names’, ASE 6 (1977), 1–13, at 10Google Scholar; Ekwall, Dictionary, s.nn. Nately and Netley Marsh.

¹²⁶ Cf. Stevenson, ‘Dr. Guest’, p. 637, n. 55; Stenton, , Preparatory to ‘Anglo-Saxon England’, pp. 278–9Google Scholar; and Gelling, Margaret, The Place-Names of Oxfordshire, EPNS 23–4 (Cambridge, 1953–1954) 1, 238–9, and 11, 444Google Scholar, and Signposts, p. 14.

¹²⁷ I do not know that this has been suggested; but cf. Chadwick's comment on the earlier apparent folk-etymologies: ‘The analogy of similar stories in other lands would lead us to infer that the personal names had been created out of the place-names. Of course one such case by itself is inconclusive…It is the uniformity of the above list which excites suspicion’ (Origin, p. 21).

¹²⁸ ibid. pp. 26–8.

¹²⁹ Asser's Life of King Alfred, ed. Stevenson, W. H. (Oxford, 1904), pp. 1–4Google Scholar (chs. 1–2).

¹³⁰ Differently Stevenson, ibid. pp. 172–5; but cf. Plummer, , Two Chronicles 11, 14Google Scholar; Chadwick, , Origin, p. 21, n. 1Google Scholar; and Campbell, , Grammar, pp. 245–7Google Scholar. (Camden's still-current equation with Carisbrooke is a modern folk-etymology.)

¹³¹ On Cerdic see Jackson, Kenneth, Language and History in Early Britain (Edinburgh, 1953), pp. 554, 557, 613–14, 653 and 689Google Scholar. The upshot is that the Old English name was borrowed after the mid-sixth-century British syncope, so that Cerdic is unlikely to have been a fifth-century figure whose name was handed down in English speech. The name comes from British *Caratīcos (from *Corotīcos only if borrowed after the seventh-century Primitive Welsh i-affection). Cynric could be a Germanic or a Celtic name; Morris is mistaken in thinking Cunorix specifically Irish (Age of Arthur, pp. 125 and 225).

¹³² Cf. Chadwick, , Origin, pp. 25–6 and 28–31Google Scholar; Stenton, , Anglo-Saxon England, pp. 25 and 69, n. 5Google Scholar; Sawyer, , From Roman Britain to Norman England, p. 50Google Scholar. I agree with Stenton that Cerdic is probably not derived from the place-names. The reverse may be true, in that they (and the attested use of Cerdic as a personal name in eighth-century Wessex, along with Cynric) may reflect the fame of Cerdic in legend.

¹³³ Hodgkin, R. H., A History of the Anglo-Saxons (Oxford, 1935) 1, 129Google Scholar; Morris, , Age of Arthur, pp. 103–4.Google Scholar

¹³⁴ Myres, , Pottery and the Settlement, pp. 117–18Google Scholar; Swanton, M. J., The Spearheads of the Anglo-Saxon Settlements (London, 1973), pp. 118, 121 and 142Google Scholar; Morris, , Age of Arthur, pp. 294–5 and 629–30.Google Scholar

¹³⁵ As Myres himself observes (p. 101).

¹³⁶ Leeds, E. T., ‘The Distribution of the Anglo-Saxon Saucer Brooch in Relation to the Battle of Bedford, a.d. 571’, Archaeologia 63 (1912), 159–202, at 173CrossRefGoogle Scholar; Farley, Michael, ‘Saxon and Medieval Walton, Aylesbury: Excavations 1973–4’, Records of Buckinghamshire 20 (1975–1978), 153–290, at 175Google Scholar. ‘Bedford’ for B(i)edcanford is speculative, Morris's, ‘near Bedford’ (Age of Arthur, p. 211) amusing.Google Scholar

¹³⁷ ibid. p. 227; Brown, G. Baldwin, The Arts in Early England (London, 1903–1937) 111, 52–3, and iv, 645 and 662Google Scholar; Brown, David, ‘Archaeological Evidence for the Anglo-Saxon Period’, Studies in the Archaeology and History of Cirencester, ed. Alan, McWhirr, BAR 30 (Oxford, 1976), 19–45, at 32–3Google Scholar. (All these are discussing Cirencester. The Chronicle is the only reason for supposing a British presence there as late as 577.) It is impossible to see how the 571 annal can be relevant, as often supposed, to a possible British enclave in the Chilterns, which are not mentioned.

¹³⁸ Stenton, , Anglo-Saxon England, pp. 27–8Google Scholar; Myres, , Pottery and the Settlement, pp. 47, 57, 115–16, and 115, n. 1.Google Scholar

¹³⁹ For criticism of the literary evidence used by Stenton (pp. 6–8) see BOH 11, 28–9; Thompson, ‘Procopius’, p. 507; and Sims-Williams, ‘Gildas’, pp. 25–7. For criticism of Myres's archaeological evidence see Hills, ‘Archaeology of Anglo-Saxon England’, p. 325; and Dickinson, Tania M., ‘On the Origin and Chronology of the Early Anglo-Saxon Disc Brooch’, ASS AH 1, BAR Brit. ser. 72 (Oxford, 1979), 39–80, at 62, n. 118.Google Scholar

¹⁴⁰ Brown, Baldwin, Arts iv, 615Google Scholar; Alcock, Leslie, Arthur's Britain, rev. ed. (Harmondsworth, 1973), p. 118Google Scholar. Cf. Evison, , Fifth-Century Invasions, pp. 7 and 84Google Scholar. It would be perverse to insist on the fact that the 571 annal is ambiguous.

¹⁴¹ The original first element of Eynsham is debated (cf. Gelling, , Place-Names of Oxfordshire 11, 259)Google Scholar, but is hardly OW Eugein (earlier Ougein) as Ekwall suggests (Dictionary, p. 172).

¹⁴² The Chronicle of Æthelweard, ed. Campbell, A. (London, 1962), p. 13Google Scholar; Monumenta Historica Britannica 1, ed. Petrie, Henry (London, 1848), p. 525Google Scholar. See Campbell, ‘Bede's Words for Places’, pp. 49–50.

¹⁴³ Sawyer, P. H., Anglo-Saxon Charters: an Annotated List and Bibliography (London, 1968)Google Scholar (hereafter cited as S), no. 217 (see also no. 887), and Cam, Helen M., Liberties and Communities in Medieval England (Cambridge, 1944), pp. 74, 76, 95 n. 1, 99, 106 and 109Google Scholar. With what follows cf. in general and in particular Anderson, Olof S., The English Hundred-Names (Lund, 1934–1939).Google Scholar

¹⁴⁴ S 1436 (Ekwall, , Dictionary, p. 172)Google Scholar; cf. Cam, , Liberties, pp. 76, 95 n. 1, 106 and 109–10Google Scholar; Gelling, , Place-Names of Oxfordshire 11, 245Google Scholar; see also S 210.

¹⁴⁵ Cam, , Liberties, pp. 83Google Scholar, n. 5, and 106 (Aylesbury), and 68, 82 n. 13 and 125–8 (Luton). Cf. Sawyer, , From Roman Britain to Norman England, pp. 199–200.Google Scholar

¹⁴⁶ Cf. ibid. pp. 86 and 88; Campbell, ‘Bede's Words’, p. 50, n. 20.

¹⁴⁷ S 1258 (cf. Whitelock, , Documents, p. 20 and no. 79)Google Scholar. For Cookham and its seven hundreds see Cam, , Liberties, pp. 82, n. 2; 98 and 106Google Scholar; Stenton, , Anglo-Saxon England, p. 301.Google Scholar

¹⁴⁸ S 1436 and 138.

¹⁴⁹ See Stenton, , Anglo-Saxon England, p. 209.Google Scholar

¹⁵⁰ Jackson's, remarks in Language and History, pp. 464–6 and 677Google Scholar, have given rise to misunderstanding; e.g. Harrison, (Framework, p. 123Google Scholar) says that ‘to a Celtic scholar they “have the look of contemporary records”’. Jackson gives no linguistic evidence for his impression. Despite a long-standing misapprehension (see Rhys, J., Celtic Britain, 3rd ed. (London, 1904), p. 138)Google Scholar, the variant -mægl in later manuscripts of the Chronicle is surely just an Old English spelling; cf. Jackson, Kenneth, ‘The British Language during the Period of the English Settlements’, Studies in Early British History, ed. Chadwick, Nora K. (Cambridge, 1954), pp. 61–82, at 77, n. 2Google Scholar. On Mægla see below, n. 177.

¹⁵¹ See Jackson, , Language and History, pp. 616–17Google Scholar (Farinmail is not cited). For Ffern- > Ffyrn- (influenced by ffyrnig?) cf. ibid. p. 665.

¹⁵² Cf. the possible Gaulish name Vlidorix ‘Feast-King’ (Evans, D. Ellis, Gaulish Personal Names (Oxford, 1967), p. 248)Google Scholar. I would prefer to assign Fernmail to the root *sper- (cf. Hamp, Eric P., ‘Varia’, Études celtiques 19 (1982), 141–2Google Scholar; Lindeman, F. O., ‘On some Compound Verbs in the Brittonic Languages’, BBCS 29 (1980–1982), 504–5).Google Scholar

¹⁵³ See Williams, Ifor, The Beginnings of Welsh Poetry, 2nd ed. (Cardiff, 1980), pp. 147–50Google Scholar; Gould, J., ‘Letocetum: the Name of the Roman Settlement at Wall, Staffs.’, South Staffordshire Archaeol. and Hist. Soc. Trans. 13 (1971–1972), 51–4Google Scholar; Jones, Glanville R. J., ‘Continuity Despite Calamity: the Heritage of Celtic Territorial Organization in England’, Jnl of Celtic Stud. 3 (1981–1982), 1–30, at 22–30Google Scholar; Gruffydd, R. Geraint, ‘“Marwnad Cynddylan”’, Bardos: Penodau ar y Traddodiad Barddol Cymreig a Cheltaidd cyflwynedig i J. E. Caerwyn Williams, ed. Gruffydd, R. Geraint (Cardiff, 1982), pp. 10–28Google Scholar. For Welsh heroes called Cynfael and Ffernfael see Jones, ‘Black Book of Carmarthen “Stanzas of the Graves”’, pp. 113, 124, 130 and 134.

¹⁵⁴ Not in 449 itself! Myres (‘The Adventus Saxonum’, p. 230, n. 14) and Hawkes (‘Jutes’, p. 93) build too much on the MS E annal for 443; cf. Harrison, , Framework, p. 120, n. 1.Google Scholar

¹⁵⁵ With Morris, (Age of Arthur, pp. 40 and 86Google Scholar) and Johnson, Stephen (Later Roman Britain (London, 1980), p. 122).Google Scholar

¹⁵⁶ Origines Celticae 11, 160. Æthelweard does not provide evidence for the antiquity of this system. It is interesting that the Kentish battles in the Historia Brittonum, ch. 44, lack any chronology.

¹⁵⁷ Sisam thought Æsc the result of miscopying Bede's Oisc, implying that the living Kentish tradition was dead (‘Anglo-Saxon Royal Genealogies’, p. 338); however, Æsc may be a modification of Kentish Esc inspired by folk-etymology (equation with OE æsc ‘ash, spear’, Kentish ěsc).

¹⁵⁸ See below, p. 37. There is endless scope for errors of a year by confusion between events being ‘in the nth year’ and events being ‘after n years’, and the formula ‘when n years had passed from Christ's birth’ is also ambiguous. These problems (which also attend my own formulations, deliberately) may lie behind the confusion about whether the landing was in 494 or 495; cf. the Regnal Table (Whitelock, , Documents, p. 146)Google Scholar, the Chronicle's 495 annal, and Æthelweard, ed. Campbell, pp. xl and 11.Google Scholar

¹⁵⁹ On the use of the Old English poetic figures thirty and fifty see Whitelock, , Documents, p. 161, n. 1Google Scholar; Harrison, , Framework, p. 132, n. 10.Google Scholar

¹⁶⁰ Chadwick thought the West Saxon dates were based on those for Hengest (Origin, pp. 22, n. 4, and 34), but the reverse is more likely since an explanation for the origin of the former can be suggested (see below).

¹⁶¹ See the table given by Kirby, ‘Problems’, p. 16. There are discrepancies between the texts, which should be cleared up by Dr Dumville's projected study of them, but Chadwick's point will undoubtedly stand, granted the point made above, n. 158.

¹⁶² Origin, p. 22, n. 4.

¹⁶³ Whitelock, , Documents, pp. 147–8.Google Scholar

¹⁶⁴ Cf. Sawyer, , From Roman Britain to Norman England, pp. 45–50Google Scholar; and, in general, Dumville, David N., ‘Kingship, Genealogies and Regnal Lists’, Early Medieval Kingship, ed. Sawyer, P. H. and Wood, I. N. (Leeds, 1977), pp. 72–104.Google Scholar

¹⁶⁵ Harrison, Kenneth, ‘Early Wessex Annals in the Anglo-Saxon Chronicle’, EHR 86 (1971), 527–33CrossRefGoogle Scholar; apparently modified in ‘The Primitive Anglo-Saxon Calendar’, Antiquity 47 (1973), 284–7CrossRefGoogle Scholar, and Framework, pp. 1–14 and 127–32. The clear cases are 514 = 495 and 527=508. 519 = [500] is uncertain. Despite Stenton, (Anglo-Saxon England, p. 20)Google Scholar, Æthelweard's 500 annal may have been taken from the Regnal Table (along with the figure 494) and not from a variant text of the Chronicle proper. An explanation of the 519 date is advanced below.

¹⁶⁶ For the text see Krusch, , Studien, p. 69Google Scholar (also Migne, Patrologia Latina 67, col. 495). The fact that Dionysius gives the 513–31 cycle according to the obsolete era of Diocletian (A.Dioc. 229–47) may have added to the confusion. In 532 he begins a.d. dating; cf. Harrison, , Framework, pp. 34–5Google Scholar. On Dionysiac tables in England see ibid. pp. 30–75.

¹⁶⁷ Why did he put 514, not 513? Perhaps he did put 513 (note the evidence for 494 beside 495) and there was a later dislocation. Or maybe the entry was spread ambiguously over 513–14 (cf. the Lindisfarne Annals (sic) reproduced by Jones, , Saints' Lives, facing p. 10).Google Scholar

¹⁶⁸ Ed. Campbell, p. 19. Campbell suggests this is an error for 140 (i.e. 494).

¹⁶⁹ Whitelock, , Documents, p. 146.Google Scholar

¹⁷⁰ Æthelweard, ed. Campbell, p. 11: ‘Sexto etiam anno aduentus eorum occidentalem circumierunt Britanniæ partem, quæ nunc Vuestsexe nuncupatur.’

¹⁷¹ ibid. p. 28: ‘a principatu Cerdic ex quo subiugauit Brittanniæ partem occidentalem, qui fuit Ecgbyrhtes regis genitor decimus, transacti sunt anni trecenti CCC’. Cf. ibid. p. 20.

¹⁷² Preparatory to ‘Anglo-Saxon England’, p. 122.

¹⁷³ HE 1.16 and 22. See above, pp. 20–1.

¹⁷⁴ ‘The Annales Cambriæ and Old-Welsh Genealogies from Harleian MS. 3859’, ed. Phillimore, Egerton, Y Cymmrodor 9 (1888), 141–83, at 159Google Scholar. Cf. Jackson, Kenneth, ‘The Site of Mount Badon’, Jnl of Celtic Stud. 2 (1953–1958), 152–5Google Scholar. Jackson wrongly restricts the field to southern England (p. 154). He revives Guest's identification with Badbury Rings (OE Baddanburg in the Chronicle s.a. 900), while admitting the impossibility of regularly deriving Baddan- from Celtic; note also that the Romano-British name was Vindocladia (Rivet and Smith, Place-Names, p. 500). Alcock rightly observes that mons in itself does not imply a hillfort (‘Her…gefeaht’, pp. 419–21).

¹⁷⁵ Ed. W. Stubbs, Rolls Ser. (London, 1887–9) 1, 23. After this William also adds the battle ‘iuxta montem qui dicitur Penne’, which is in the Chronicle s.a. 658 (Cenwalh's only specifically anti-British battle there); there is no reason to suppose that William gives the two battles in chronological order. The battle of Wirtgernesburg need not be identified with the battle of Bradford-on-Avon of 652, which Æthelweard calls bellum ciuile and which was probably missing in the E-type Chronicle used by William (see Whitelock, , Documents, pp. 120–1 and 164, n. 5).Google Scholar

¹⁷⁶ Alcock, , Arthur's Britain, p. 117.Google Scholar

¹⁷⁷ Cunliffe, Barry, Excavations at Portchester Castle (London, 1975–1977) 11, 2Google Scholar. The annal runs: ‘In this year Port and his two sons Bieda and Maegla came to Britain with two ships at the place which is called Portsmouth; and there they killed a [young] British man of very high rank.’ SwEETSuggested that the last phrase (‘ofslogon anne giongne Brettisc monnan, swiþe æþelne monnan’, in MS A) derived from poetry (‘Some of the Sources’, p. 312). Alternatively a name (an eponym like Natanleod?) may have dropped out, or Mægla, who is not named as Port's son by Æthelweard, could be the name of the Briton, displaced in the extant Chronicle. Note, however, that Mǣgla may well be a genuine Old English name unrelated to Celtic Maglo-; see Stevenson, ‘Beginnings of Wessex’, pp. 36–7.

¹⁷⁸ ibid. p. 35; Stenton, , Anglo-Saxon England, p. 20, n. 1Google Scholar; Hodgkin, , History of the Anglo-Saxons 1, 127.Google Scholar

¹⁷⁹ Finberg, H. P. R., ‘Roman and Saxon Withington’, Lucerna (London, 1964), pp. 54–5.Google Scholar

¹⁸⁰ BOH 11, 28.

¹⁸¹ Saxons in England 1, 22.