^{page 121 note 1} I am grateful to the Dean and Chapter of Canterbury, and particularly their archivist Miss Anne M. Oakley, M.A., fot the help they have given. For any usefulness which may attach to the first and second parts of this paper, the credit must go respectively to Dr William Urry and Professor E. J. Dobson. For the discovery of the manuscript, see Ker, N. R., Catalogue of Manuscripts containing Anglo-Saxon (Oxford, 1957), p. vii.Google Scholar

^{page 121 note 2} Ibid. 96.

^{page 121 note 3} References are by page and line to Bischofs Wærferth von Worcester Übersetzung der Dialoge Gregors des Grossen, ed. Hecht, H., Bibliothek der angelsachsischen Prosa 5 (Leipzig, 1900Google Scholar; repr. Darmstadt, 1965), based on the other manuscripts.

^{page 121 note 4} Loyn, H. R., the only other person, so far as I know, who offers a date for this manuscript, says it ‘cannot be dated before the second quarter of the eleventh century’ (‘The Term Ealdorman in the Translations Prepared at the Time of King Alfred’, EHR 68 (1955), 519).Google Scholar

^{page 121 note 5} The library numbers the bifolium made up of fols. 2 and 3 as ‘(i)’ and the bifolium made up of fols. 1 and 4 as ‘(ii)’.

^{page 130 note 1} Almost every modern scholar has applied ‘C’, ‘O’ and ‘H’ to the Corpus, Cotton and Hatton manuscripts respectively. Here ‘A’ signifies the Canterbury fragment, because of its textual importance explained below.

^{page 130 note 2} There are also two Old English glosses in Canterbury Cathedral Library, Add. 32, a fragment of bk III of the Latin Dialogues, which was discovered, like A, by Dr Urry. Ker (Catalogue, p. lxiii) has described and printed the glosses (ref010 for the Latin are to the edition of U. Moricca (Rome, 1924) and for the Old English to Hecht's edition; I have checked the readings against the manuscript; I use the conventions for damaged text explained above, pp. 122–3): ir 164.16 non urguerent, 2nd u crossed out by a later band OE gloss ne nyddan 200.5 CO ne nyddon ir 164.21 <defectus> OE gloss wanunge 200.10 C warninge O wanunge Ker (Ibid.) both the Latin and the Old English to s. xi in., a time subsequent to the writing of MS A. If these glosses were taken from A, that manuscript once contained bk in of Werferth's translation. But there are other possibilities. The glosses could have come from another, lost manuscript (C and O are too late; see below), or they could have been completely independent of Werferth's translation.

^{page 131 note 1} For the existence of two versions, see, e.g., Johnson, H., Gab es zwei von einander unabhängige altenglische Übersetzungen der Dialoge Gregors? (Ph.D. thesis, Berlin, 1884)Google Scholar. In the Einleilmg (Hamburg, 1907) to his edition, Hecht expounded the relation of C, O and H (pp. 29–35), drew two variations of a stemma for them (p. 35) and further demonstrated that H witnesses the revision (pp. 130–4). K. Sisam subsequently discredited this stemma, but not Hecht's conclusion about the relation of the two versions to each other; see ‘An Old English Translation of a Letter from Wynfrith to Eadburga (A.D. 716–17) in Cotton MS. Otho C I’Google Scholar; and ‘Addendum: The Verses Pref018 to Gregory's Dialogues’, Studies in the History of Old English Literature (Oxford, 1953), pp. 200–3 and 229 and n. I.Google Scholar

^{page 131 note 2} Catalogue, p. 106.

^{page 131 note 3} Ibid. 236.

^{page 132 note 1} Example 9 involves both CO/A and AO/C.

^{page 133 note 1} ‘The Text of the Old English Translation of Gregory's Dialogues’, Neophilogus 22 (1937), 293–4, where Harting quoted no less than thirteen shared errors. In the same paper Harting went on to say that ‘there may have been one or more intermediate stages between O and C od Wasrferð's copy, although I believe it is impossible to prove this’ (p. 287). Hecht, on the other hand, called the exemplar common to C and O ‘W’, by which he meant Werferth (Einkitung, p. 35).Google Scholar

From a comparison of C and O with the Latin text and variants available in Moricca's edition, Harting had also concluded that ‘a later scribe, whether C or O or another, ref026 occasionally to a Latin text of the Dialogues differing in particulars from the one Waaferð used’ (pp. 287–8). The reading Separates AO, Reparatus C (at Ir 12 (307.24), (Ir 15 (307.26) and Ir 24 (308.2)) strongly suggests that it is the scribe of C or of an exemplar between β and C who independently consulted a Latin manuscript: all the manuscripts (a total of about sixty) and all the printed editions (about forty) of the Latin text of the Dialogues which I have consulted read Reparatus, or occasionally Paratus, never Separatus. (The manuscripts include all those in the British Library and Cambridge University, Bodleian and Oxford and Cambridge college libraries, as well as Heref027, Cathedral Library, O. I. x; London, Lambeth Palace Library, 204; Rouen, Bibliothèque Municipale, A. 337 (506); Salisbury, Cathedral Library, 95 and 96; Sankt Gallen, Stiftsbibliothek, 213; and Würzburg, Universitätsbibliothek, M.p.th. f. 19. The printed editions include all those in the libraries mentioned first above.) Apart from MSS A and O of Werferth's translation, I have observed Separatus only in the Old Norse translation of the Dialogues (printed by Unger, C. in Heilagra Manna Sogur I (Christiania, 1877)Google Scholar and by Bjarnarson, þ. in Leifar fornra kristinna fræa islenzkra … (Copenhagen, 1878).Google Scholar