^{page 51 note 1} The Homilies of the Anglo-Saxon Church: the First Part containing the Sermones Catholici or Homilies of Ælfric, ed. B. Thorpe (London, 1844–1846) 1, 99Google Scholar. I am grateful to Professor Dorothy Whitelock for this reference and for much else, as will appear.

^{page 51 note 2} Byrhtferth's Manual, ed. S. J. Crawford, Early English Text Society 177 (London, 1929), 63Google Scholar. The Menologium, of about the same date, after mentioning the Nativity goes on to say that January is the start of the year: Anglo-Saxon Minor Poems, ed. E. van K. Dobbie (New York, 1942), p. 49.Google Scholar

^{page 52 note 1} Bedae Opera de Temporibus, ed. C. W. Jones (Medieval Academy of America, Cambridge, Mass., 1943), P. 73.Google Scholar

^{page 52 note 2} Venerabilis Baedae Opera Historica, ed. C. Plummer (Oxford 1896)Google Scholar. Cited henceforth as HE, and notes as Plummer, Opera Historica 11.

^{page 52 note 3} Krusch, B., ‘Studien zur christlich mittelalterlichen Chronologie’, Abhandlungen der Preussischen Akademie der Wissenschaften, Phil.-Hist. Klasse 7, no. 8 (Berlin, 1938), 70Google Scholar; Migne, Patrologia Latina 67, cols. 493–8.

^{page 52 note 4} Krusch, , ‘Studien’, p. 60Google Scholar and Ruhl, F., Chronologic des Mittelalters und der Neuzeit (Berlin, 1897), pp. 179–80Google Scholar. Only in the Alexandrian cycle did the pattern of epacts change in September. Dionysius had to face the problem of reconciling Alexandrian (luni-solar) epacts with the Roman (solar) calendar.

^{page 53 note 1} Jones, , Opera de Temporibus, p. 211.Google Scholar

^{page 53 note 2} Harrison, K., ‘Early Wessex Annals in the Anglo-Saxon Chronicle’, EHR 86 (1971), 527–33.CrossRefGoogle Scholar

^{page 53 note 3} Jones, , Opera de Temporibus, p. 223.Google Scholar

^{page 53 note 4} Poole, R. L., Studies in Chronology and History (Oxford, 1934), p. 8.Google Scholar

^{page 53 note 5} A point already noticed by Jones, C. W., Saints' Lives and Chronicles in Early England, (Ithaca, N. Y., 1947), p. 210Google Scholar, n. 65.

^{page 53 note 6} Jones, , Opera de Temporibus, pp. 236 and 298Google Scholar. In his notes pp. 354–7, Jones would partly attribute the preservation of 1 January as New Year's Day to Bede and his followers.

^{page 54 note 1} Monumenta Germaniae Historica, Epistolae 111, ed. E. Dümmler (Berlin, 1892), 301.

^{page 54 note 2} Jaffé, P., Regesta Pontificum Romanorum, 2nd ed. (Leipzig, 1885), p. ix.Google Scholar

^{page 54 note 3} Jones, , Opera de Temporibus, p. 268Google Scholar. Dr N. Brooks has drawn my attention to a paper by Halkin, J., ‘La Nouvelle Année au 23 Septembre’, AB 90 (1972), 56Google Scholar, where it is stated that, from the fourth century until the middle of the fifth, the Byzantine year was sometimes reckoned from the birthday of the Emperor Augustus, 23 September, coinciding with the autumnal equinox. Yet the choice of 24, rather than 22 or 23, September is curious. In a calendar so adjusted that epacta nulla falls on 1 January, as in the first year of a Dionysiac nineteen-year cycle, there is another epacta nulla on 24 September, which may explain Bede's preference for this day; the equinox can vary from year to year, whereas the epact is fixed.

^{page 55 note 1} Plummer, , Opera Historica 11, 211.Google Scholar

^{page 55 note 2} The Life of Bishop Wilfrid, by Eddius Stephanus, ed. B. Colgrave (Cambridge, 1927) ch. LIX, p. 128Google Scholar. Bede, who was not writing a political history, does not mention the affair.

^{page 55 note 3} Harrison, K., ‘The Reign of King Ecgfrith of Northumbria’, Yorkshire Archaeol. Jnl 44 (1972), 79.Google Scholar

^{page 55 note 4} Poole, , Studies, p 40.Google Scholar

^{page 56 note 1} Poole, , Studies, p. 43.Google Scholar

^{page 56 note 2} England and the Continent in the Eighth Century (Oxford, 1946), p. 267.Google Scholar

^{page 56 note 3} De Natura Rerum, PL 90, col. 244.

^{page 56 note 4} According to ASC C, the comet of 1066 (Halley's) appeared on 24 April and ‘swa scean ealle þa. vii. niht’ – a spectacle of shortest duration, yet it had done its work and the Bayeux Tapestry makes it foreshadow the Battle of Hastings in October.

^{page 56 note 5} Harrison, ‘The Reign of King Ecgfrith’, p. 79.

^{page 56 note 6} HE rv. 5.

^{page 56 note 7} That Theodore used the Greek indiction is shown by the date ‘sub die XV Kalendas Octobres, indictione VIIIa’, 17 September 679, for the Council of Hatfield (HE rv.15) in Ecgfrith's tenth year. The comet was dated by Bede to August 678, in Ecgfrith's eighth year; hence he came to the throne between 1 and 17 September 670.

^{page 56 note 8} Levison, , England and the Continent, p. 267.Google Scholar

^{page 57 note 1} It seems that the Dionysiac tables were formally introduced at the Synod of Whitby in 664 (Poole, Studies, p. 32), but how much earlier they came into use is not at the moment clear.

^{page 57 note 2} Harrison, ‘Reign of King Ecgfrith’, p. 81.

^{page 57 note 3} Jones, , Opera de Temporibus, p. 269Google Scholar. In practice he could get the number from a Dionysiac table.

^{page 57 note 4} Levison, (England and the Continent, p. 277Google Scholar), in arguing that Bede began the year at Christmas, quotes a passage from De Temporum Ratione, ch. XLVII (Jones, , Opera de Temporibus, p. 267Google Scholar), in which the occurrence of that festival at Rome is dated ‘anno ab eius incarnatione iuxta Dionysium septi-ngentesimo primo, indictione quarta decima’. Levison continues: ‘The 14th Indiction began on 1 September 700; therefore Christmas 701, as mentioned, would have been Christmas of A.D. 700 if the number of the year had changed on the 1st of January.’ It was indeed Christmas 700, because the year A.D. 701 is here qualified by the indiction.

^{page 57 note 5} HE 11. 20.

^{page 57 note 6} HE 111. 9.

^{page 57 note 7} HE v. 24.

^{page 57 note 8} HE 111. 1 and 9.

^{page 57 note 9} Handbook of British Chronology, ed. Sir F. Maurice Powicke and E. B. Fryde, 2nd ed. (Royal Historical Society, London, 1961), p. 11.Google Scholar

^{page 57 note 10} HE 111. 9–13 and iv. 14.

^{page 57 note 11} HE 111. 13. In contrast, Bede does not state the day or month of Wilfrid's death, and gives the year only by inference: for a discussion see my forthcoming article in the Yorkshire Archaeol. Jnl, ‘The Deaths of King Aldfrith and of Bishop Wilfrid’. Some difficulties over the career of Archbishop Deusdedit seem to have been resolved by Grosjean, P., ‘La Date du Colloque de Whitby’, AB 78 (1960), 235–8Google Scholar. Bede's date for the consecration of Archbishop Willibrord is considered below, p. 69.

^{page 58 note 1} HE 11. 18.

^{page 58 note 2} HE 11. 17.

^{page 58 note 3} Plummer, , Opera Historica 11, 110.Google Scholar

^{page 58 note 4} The Anglo-Saxon Chronicle: a Revised Translation, ed. D. Whitelock, D. C. Douglas and S. I. Tucker, (London, 1961), p. 116Google Scholar. (This work is cited henceforth as Whitelock, ASC.) The C text is here starting the year at Lady Day; Easter was on 31 March; and by dating Robert's appointment to mid-Lent 1050, and his return to 1051, with the information from E that he went to Rome ‘in the course of the same Lent’, C tells us that he started before 25 March.

^{page 58 note 5} Poole, R. L., ‘The Early Correspondence of John of Salisbury’, Proc. of the Brit. Acad. 11 (1924) 31–2Google Scholar; repr. Studies, pp. 263–4. I am grateful to Professor Whitelock for raising this point of time.

^{page 58 note 6} HE iv. 1.

^{page 58 note 7} Although in his formal letter, 5 November 734, to the bishop (later archbishop) of York, he employed the formal style of the indiction; see Plummer, , Opera Historica 1, 423Google Scholar and 11, 388.

^{page 58 note 8} HE 11. 4.

^{page 59 note 1} HE v. 24.

^{page 59 note 2} Simeonis Monachi Opera, ed. T. Arnold, Rolls Series (1885).

^{page 59 note 3} Celt and Saxon: Studies in the Early British Border, ed. N. K. Chadwick (Cambridge, 1963), p. 63.Google Scholar

^{page 59 note 4} Hart, C., ‘The Ramsey Computus’, EHR 85 (1970), 29.CrossRefGoogle Scholar

^{page 59 note 5} Whitelock, , ASC p. xiv.Google Scholar

^{page 59 note 6} Harrison, K., ‘The Beginning of the Year among Bede's Successors’, Yorkshire Archaeol. Jnl 42 (1967), 193Google Scholar. The opening paragraphs of this paper are now obsolete.

^{page 59 note 7} MGH, Epistolae iv, ed. E.Dümmler (Berlin, 1892), 231; Levison, , England and the Continent, p. 277.Google Scholar

^{page 59 note 8} Printed, Cartularium Saxonicum, ed. Birch, W. de G. (London, 1885–1893)Google Scholar, cited henceforth as BCS. This discussion would not have been possible without the help of Professor Whitelock, who supplied me with a select list of some eighty documents and, more important still, her comments on the reliability of the material and on a number of debatable points. I have also drawn on Sawyer, P. H., Anglo-Saxon Charters: an Annotated List and Bibliography (Royal Historical Society, London, 1968)Google Scholar, and occasionally on M. Treiter, ‘Die Urkundendatierung in angel-sächsischer Zeit’, Archiv für Urkundenforschung 7 (1921), 53.Google Scholar

^{page 59 note 9} Whitelock, , ASC pp. 31 and 36.Google Scholar

^{page 60 note 1} Councils and Ecclesiastical Documents relating to Great Britain and Ireland, ed. A. W. Haddan and W. Stubbs (Oxford, 1869–1878) 111, 559Google Scholar. As a parallel to this use of solium, Alcuin writes to Æthelred of Northumbria (790–5) ‘non decet te in solio sedentem regni rusticis uiuere moribus’ (MGH, Epistolae rv, 71).

^{page 60 note 2} The Martyrology of Oengus, ed W. Stokes, Henry Bradshaw Society 29 (London, 1905), 72Google Scholar and The Calendar of St Willibrord, ed. H. A. Wilson, HBS 55 (London, 1918), 35Google Scholar; on the whole question see Holweck, F. G., Calendarium Liturgicum (Philadelphia, 1925), pp. 230 and 258.Google Scholar

^{page 60 note 3} Haddan, and Stubbs, , Councils 111, 468Google Scholar. Professor Whitelock thinks that sedenti in solio could merely mean ‘occupying the archiepiscopal see’. Dr N. Brooks takes it as evidence that Wulfred was not yet consecrated, and has kindly furnished me (9 June 72) with the following note, which corrects my earlier view that the consecration took place before 26 July 805:

‘If one accepts BCS 322 as evidence that Wulfred was not consecrated by 26 July 805, then the other Wulfred charters fall into place. One can accept the evidence of BCS 335 that his consecration must have been after 1 August 805, and of BCS 378 that the Council of Clofeshob of 824 was in his nineteenth year; BCS 379 shows this to have taken place at the end of October, and Wulfred's consecration is therefore pushed to October 805 or later. Thus there is no need to amend BCS 355, and no need to assume that there was one Council at Clofeshob before 24 July and another in October [which I had previously thought]. In this context there is no significance in the fact that BCS 378 and 379 are in different formulas. BCS 378 is in favour of Canterbury, and drafted by the Canterbury scriptorium, BCS 379 in favour of Worcester and drafted by that scriptorium.’

This straightforward view has the advantage of avoiding emendation. The former charter, witnessed by Beonna (of Hereford) as electus, will have been drafted earlier than the latter, where he is episcopus. Perhaps sedens in solio can be taken in the sense of ‘administering the see’.

^{page 61 note 1} Whitelock, , ASC pp. xii–xvii.Google Scholar

^{page 61 note 2} Historia Monasterii S. Augustini Cantuariensis, ed. C. Hardwick, RS (1858), pp. 13 and 339.

^{page 61 note 3} The forged BCS 338 and suspicious BCS 349, with no regnal year in any case, have been omitted.

^{page 62 note 1} Expressed in another way, the charters in tables 2 and 3 dated in the critical period 25 September to 24 December (or 30 December), that is, BCS 308, 309, 310, 312, 340, 348, 350 and 379, carry an indictional number that is one too few if we assume that the people who drafted them were using an indiction beginning in September and a Year of Grace beginning on 25 December (or 1 January). Hence the ecclesiastical scriptoria began their indictional years and Years of Grace at the same time. (It should be noted, as Dr Brooks reminds me, that BCS 310 and 312 are contemporary copies.) On the evidence of BCS 350 the year began at Christmas, in agreement with other evidence from Alcuin and the Northumbrian annals. Of Carolingian practice it has been said that ‘fréquemment on a fait concorder l'indiction avec l'année de l'Incarnation ou même avec l'année du règne’ (Giry, A., Manuel de Diplomatique (Paris, 1894), p. 728).Google Scholar

^{page 62 note 2} In a paper by P. Chaplais, to which Professor Whitelock has drawn my attention, ‘Some Early Anglo-Saxon Diplomas on Single Sheets: Originals or Copies ?’ Jnl of the Soc. of Archivists 3 (1965–1969), 325Google Scholar, it is stated that this charter shows use of the ‘Bedan indiction’ of 24 September. This would be true only if we could be assured that the scribe did not take the indiction from the column adjacent to the annus domini in a Dionysiac table, as indicated above.

^{page 62 note 3} Unfortunately the reliable BCS 202–4, 223, 230 and 239 do not carry regnal years, and BCS 232, 256, 265, 267 and 269 are too muddled to be of use.

^{page 62 note 4} HE v. 23 and 24.

^{page 63 note 1} HE v. 24.

^{page 63 note 2} Northumbrian annals in Simeonis Monachi Opera 11, 32. Dates from this source are usually reliable.

^{page 63 note 3} ‘The Supremacy of the Mercian Kings’, EHR 33 (1918), 439.Google Scholar

^{page 63 note 4} Thus, apart from others, of the charters preserved in Elmham's Historia, ed. Hardwick, genuine or spurious, and ranging from 675 to 737, printed by Birch as BCS 35, 36, 42, 67, 73, 88, 90, 141, 149 and 150, all except BCS 42 are dated only by the indiction.

^{page 63 note 5} MGH, Legum 11, Capitularia Regum Francorum, ed. A. Boretius (Hanover, 1883), p. 84.Google Scholar

^{page 63 note 6} Krusch, B. in Mélanges Chatelain (Paris, 1910), p. 232Google Scholar. At the Council of Chelsea in 816 (Haddan and Stubbs, Councils 111, 593) the English bishops were encouraged to use AD because ‘qualis annus Domini conputatur, aut a quali Archiepiscopo … constitutum est illud iudicium. I owe this reference to Professor Whitelock.

^{page 64 note 1} The exceptional BCS 296, of 799, has been included in table 3 only on this account, and not for its reliability, which has been discussed by Levison, England and the Continent, pp. 230 and 249, n. 3.

^{page 64 note 2} Stenton, F. M., Anglo-Saxon England, 3rd ed. (Oxford, 1971), p. 230.Google Scholar

^{page 64 note 3} A century later Edward the Elder died on 17 July 924, but Athelstan's ‘coronation’ did not take place until 4 September 925: Whitelock, ASC, p. 69.

^{page 64 note 4} See above, p. 60, n. 3.

^{page 64 note 5} Thorogood, A. J., ‘The Anglo-Saxon Chronicle in the Reign of Ecgberht’, EHR 48 (1933), 358.Google Scholar

^{page 65 note 1} Whitelock, , ASC, p. 40Google Scholar, n. 5.

^{page 65 note 2} This date could be an argument in favour of a year beginning in September; but the change of indiction could also arise from thoughtless copying in accordance with the change of year at Christmas.

^{page 65 note 3} printed, AlsoAnglo-Saxon Charters, ed. Robertson, A. J. (Cambridge, 1939), pp. 16–19.Google Scholar

^{page 66 note 1} It will be noticed that BCS 308–12 and 379, where the evidence for Christmas dating seems to be unequivocal, were issued at Councils of Clofeshoh and must surely represent official practice.

^{page 66 note 2} Thorogood, ‘The Anglo-Saxon Chronicle in the Reign of Ecgberht’, pp. 555–7. And the annal for 827 (recte 829) opens with a lunar eclipse ‘on middes wintres mæsse niht’ – actually the early hours of Christmas Day.

^{page 66 note 3} Beaven, M. L. R., ‘The Beginning of the Year in the Alftedian Chronicle’, EHR 33 (1918), 238Google Scholar and Whitelock, , ASC p. xxiv.Google Scholar

^{page 67 note 1} A recent edition is that of Grat, F.et al. (Société de l'Histoire de France, Paris, 1964).Google Scholar

^{page 67 note 2} Stenton, F. M., The Latin Charters of the Anglo-Saxon Period (Oxford, 1955), pp. 39–43Google Scholar and Whitelock, D., ‘The Old English Bede’, Proc. of the Brit. Acad. 48 (1962), 74–5Google Scholar. In this paper (p. 66) Professor Whitelock also observes that the translator has omitted some references to the indiction.

^{page 67 note 3} So ASC, but other evidence suggests wintering in Kent some fifty years earlier; see Brooks, N., England before the Conquest: Studies in Primary Sources presented to Dorothy Whitelock, ed. Peter, Clemoes and Kathleen, Hughes (Cambridge, 1971), pp. 79–80Google Scholar. I am grateful to Mr Hunter Blair for drawing my attention to this point.

^{page 67 note 4} Whitelock, , ASC p. xxivGoogle Scholar; for a more detailed discussion see Professor Whitelock's Appendix to the introduction to Two of the Saxon Chronicles Parallel, ed. C. Plummer on the basis of an ed. by J. Earle (repr. Oxford, 1952), pp. cxxxix f.Google Scholar

^{page 68 note 1} ‘The South-Western Element in the Old English Chronicle’, Essays in Medieval History presented to T.F. Tout, ed. A. G. Little and F. M. Powicke (Manchester, 1925), p. 15Google Scholar. Two bishops of Sherborne figure in a military capacity: Ealhstan fighting the Danes in 845 and his successor Heahmund killed in battle in 871.

^{page 68 note 2} ‘The Date of King Alfred's Death’, EHR 13 (1898), 71.Google Scholar

^{page 68 note 3} ‘The Regnal Dates of Alfred, Edward the Elder and Athelstan’, EHR 32 (1917), 517.Google Scholar

^{page 68 note 4} Whitelock in Earle, and Plummer, , Two Saxon Chronicles, p. cxlii.Google Scholar

^{page 68 note 5} Above, p. 62.

^{page 68 note 6} Sawyer, , Anglo-Saxon Charters, pp. 207 ff.Google Scholar Some at least are not suspicious.

^{page 68 note 7} Memorials of St Dunstan, ed. W. Stubbs, RS (1874), p. lxxxviii. It has been calculated (Whitelock, ASC p. 5, n. 1), that Eadwig was probably crowned on 26 January, the third Sunday after Epiphany in 956 – not the second, as Stubbs has written, probably by a slip of the pen.

^{page 69 note 1} HE v. 11.

^{page 69 note 2} Wilson, , Calendar of St Willibrord, pp. 13 and 42Google Scholar and pl. xi.

^{page 69 note 3} Poole, , Studies, p. 36.Google Scholar

^{page 69 note 4} The Moore Bede, ed. P. Hunter Blair, EEMF 9 (Copenhagen, 1959), 30.Google Scholar

^{page 69 note 5} HE Praefatio.

^{page 69 note 6} HE 111. 13.

^{page 69 note 7} 725 (Wednesday), 726 (Thursday), 727 (Friday), 729 (Monday), 730 (Tuesday) and 731 (Wednesday). Willibrord was about seventy at this time and could not know he would live for another decade.

^{page 70 note 1} See above, p. 57.

^{page 70 note 2} Levison, , England and the Continent, p. 90Google Scholar, n. 2.

^{page 70 note 3} Plummer, , Opera Historica II, 61.Google Scholar