¹ The problem is an old one, but interest in it has been particularly keen since 1950; see below: Girvan, 1957 (n. 19); Gneuss, 1955 (n. 25); Sisam, 1953, 1956 (nn. 16, 20).

² For a fuller account of this view and a discussion of the conventional Canterbury origin of the MS, see my article, “From Canterbury to Lichfield,” Speculum, xxiii (Oct. 1948), 591–629.

³ To the best of my knowledge, the second phase has been undertaken only by Karl Wildhagen, “Studien zum Psal-terium Romanum in England und zu seinen Glossierungen,” Festschrift für Lorenz Morsbach, Studien zur englischen Philologie, L (Halle, 1913), 436–438. For an examination of his evidence, see my article in Speculum, xxiii, 605–607.

⁴ Festschrift, p. 436.

⁶ A case in point is the Ayenbite of Inwyt, almost certainly a translator's autograph, which is full of errors of every kind. For lengthy but far from exhaustive lists of these errors, see J. K. Wallenberg, The Vocabulary of Dan Michel's Ayenbite oflnwyt (Uppsala, 1923), pp. 317–332; Richard Morris, Dan Michel's Ayenbite of Inwyt, EETS, O.S., xxiii (London, 1866; repr. 1888), 272–279; Hermann Varnhagen, “Beiträge zur Erklärung und Textkritik von Dan Michel's Ayenbite of Inwyt,” Engl. Studien, i (1877), 379–423; ibid., ii (1879), 27–59; R. W. Evers, Beiträge zur Erklärung und Textkritik… (Erlangen, 1887), pp. 70–113; M. Konrath, “Die lateinische Quelle zu Ayenbite… und zu Sawles Warde,” Engl. Studien, xii (1889), 459–463.

⁶ Psalm and verse references are to Henry Sweet, The Oldest English Texts, EETS, O.S., LXXXIII (London, 1885), 188420—hereafter cited as OET. Sweet used modern Vulgate verse numbers; numbers in parentheses represent the verse divisions of the MS. OE readings not in Sweet's edition will nearly always be found in Ruby Roberts, “A new Collation of the Vespasian Psalter and Hymns,” Leeds Studies in English and Kindred Languages, i (1932), 22–23; or in Kuhn, “The Gloss of the Vespasian Psalter: Another Collation,” JEGP, xl (July 1941), 344–347. Latin readings not found in Sweet will soon be available, I hope, in my new edition of the text.

⁷ I have not included rechtwisnisse 44, 6 (possibly regarded by the glossator as a correct, though archaic, variant spelling) or rehlwisnis 71, 4 (a different type of error, i.e., nom. sg. for dat. sg.). Reht, shorter than rehtwisnis and requiring more concentration because of its more complicated inflection, is written 26 times, without error. Rehtheort (3 times) and reht-lic (once) are also without error. Rehtwis appears correctly 57 times, incorrectly twice: rehtwisse 117, 19 and rehtwisre (? fem. gen. sg. for gen. pi.) Hy. 8, 16.

⁸ I have checked only the readily available sources, but these are extensive enough to assure us that the readings in question, if they ever existed, were exceedingly rare and, therefore, unlikely to have been in an English MS older than the Vespasian Psalter. The chief Latin sources examined for this and other references in this article are: (1) the texts of Jerome's first, or Romana, version of the Psalter and of his second, or Gallicana, version in Migne's Patrologia Latina, XXIX, cols. 119–398, with MS variants in the footnotes; (2) the text of Jerome's last, or Hebraica, version, ibid., xxvin, 1127–1240; (3) the Psalter of Paris Bibliothèque Nationale Lat. 11947, as edited by P. Sabatier, Bibliorum Sacrorum Latinœ Versiones Antiquæ (Rheims, 1743), II, 9–293, with variants from other MSS in footnotes; (4) the Psalter in Codex Casinensis 557, as printed in A. M. Amelli, Liber Psal-morum Iuxta Antiquissimam Latinam Versionem, in Collectanea Biblica Lalina, i (Rome, 1912); (5) a critical edition of the Romana in Robert Weber, Le psautier romain, ibid., x (Rome, 1953); and (6) the variants in Wildhagen, Der Cambridge Psalter, in Bibliothek der angelsachsischen Prosa, VII (Hamburg, 1910).

⁹ The glossator usually followed the Latin word order, even when the genitive was a pronoun; e.g., 21, IS (12) heorte min (cor meum), 9, 32 (31) onsiene his (faciem suam).

¹⁰ Sweet, OET, p. 174, or A. B. Kuypers, The Prayer Book of Aedeluald the Bishop (Cambridge, 1902), p. 3.

¹¹ Eduard Brenner, Der allenglische Junius-P Salter, in Anglistische Forschungen, xxiii (Heidelberg, 1908).

¹² Die Sprache des kentischen Psalters (Halle, 1881), pp. 41, 44–45.

¹³ For discussion and examples, see Sweet, OET, p. 186.

¹⁴ Historische Grammatik der englischen Sprache (Leipzig, 1921), p. 266.

¹⁵ Speculum, xxiii, 602–604.

¹⁶ “Cynewulf and his Poetry,” Proceedings of the British Academy 1932–1933, xviii, 325–326; repr. in Sisam, Studies in the History of Old English Literature (Oxford, 1953), p. 4.

¹⁷ Wildhagen, Festschrift, p. 440. Wildhagen's opinion that certain additions in the Vespasian MS were taken directly from Regius (generally recognized as a Winchester book) has never been seriously challenged. The presence of Vespasian in Winchester is further indicated by the fact that its gloss was directly copied by the scribe of the Junius Psalter; this

¹⁸ My observations, based upon MLAA Rotograph No. 332, Brit. Mus. MS. Cotton Vespasian A. 1, bave been verified by T. J. Brown from the MS.

¹⁹ R. Girvan, who rejects Sisam's view, observes that “f is not seldom written for v (u) in documents of the early ninth century,” rev. of Gneuss, Lehnbildungen (see n. 25, below), in RES, N.S. vm (Feb. 1957), 49. The same may be said of the Latin in 8th-century texts; e.g., the Corpus Glossary, written as a tool for scholars and translators some 50–100 years earlier than the Vespasian gloss, has for ax for vorax, fugitifa-rius for fugitivarius, etc.; for further examples, see J. H. Hes-sels, An Eighth-Century Latin-Anglo-Saxon Glossary (Cambridge, 1890), p. xxvi. Corpus appears to have been written in the same dialect areas as the Vespasian gloss; see Kuhn, “The Dialect of the Corpus Glossary,” PMLA, liv (March 1939), 8–19. The 13th-century press mark of St. Augustine's Abbey, Canterbury, indicates little as to the provenience of this 8th-century MS.

²⁰ “Canterbury, Lichfield, and the Vespasian Psalter,” RES, N.S. vii(April 1956), 127.

²¹ I am not at all sure that Sisam meant to imply a causal relationship between the Durham Ritual gloss and the Vespasian gloss; very likely he did not.

²² For further examples see Sweet, OET, p. 187.

²³ The last is edited by Fritz Roeder, Der allenglische Regius-Psalter, Studien zur englischen Philologie, XVIII (Halle, 1904).

²⁴ Sacrorum Bibliorum Vulgatæ Editionis Concordantltz Hugonis Cardinalis… Recensitæ atque Emendalœ (Venice, 1768).

²⁵ Lehnbildungen und Lehnbedeutungen im Altenglischen (Berlin, 1955), p. 138.

²⁶ The last is edited by Fred Harsley, Eadwine's Canterbury Psalter, Pt. ii, EETS, O.S., xcii (London, 1889).

²⁷ Die Handschrift Junius 27 der Bibliolheca Bodleiana, Mémoires de la Société Néo-Philologique, iii (Helsinki, 1901), 43–48.

²⁸ Junius-Psailer, pp. xiii-xv; also pp. xv-xxxiii.

²⁹ Studien zu altenglischen Psailerglossen, in Bonner Beiträge zur Anglistik, xiii (Bonn, 1904), 98–102.

³⁰ Cambridger Psalter, pp. xiv-xv.

³¹ Psalterglossen, pp. 102–103, 122–123.

³² Festschrift, pp. 448–451. See also his remarks in a review of Roeder's Regius Psalter in Archiv für das Studium der neueren Sprachen und Literaturen, cxvi (1906), 157–163, and in “Zum Eadwine- und Regius-Psalter,” Engl. Studien, xxxix (1908), 196–197.

³³ Archiv, cxvi, 159; Engl. Studien, xxxix, 191.

³⁴ For a full and exceedingly useful analysis of E, see Wildhagen, Der Psalter des Eadwine von Canterbury, Studien zur englischen Philologie, xiii (Halle, 1905). Unless otherwise indicated, further references to Wildhagen will be to this work.

³⁵ Rev. of Wildhagen's Psalter des Ead. in Archiv, cxvi, 163–167.

³⁶ Gedest (facies); according to Harsley, the -t may be original rather than a corrector's work, and the ge- may have been erased. Widtihx ‘takest away, withdrawest’ is clearly a present form of widteon, although it glosses a perfect (abstraxisti).

³⁷ I use Wildhagen's designation.

³⁸ Wildhagen invoked this principle only in connection with copying from the Urform; he apparently assumed that the 12th-century scribes could copy all manner of rare and atypical forms. There is, moreover, the possibility that a copyist might generalize the endings of other classes of verbs to the 2 groups mentioned. If we suppose the Urform to have been derived from A, it will be relevant to note that, for the pr. 2 sg. of verbs of all classes, A has 85 percent s-endings. E also has a sprinkling of s-endings for verbs of all classes.

³⁹ The spelling oe is overwhelmingly predominant in the early glossaries; see Ferdinand Dieter, Über Sprache und Mundart der ältesten englischen Denkmäler (Göttingen, 1885), pp. 28–29. One of Dieter's two examples of ē (from õē) for Corpus is not a true example, i.e., cecil. In A, there are only 8 examples of ē for õē, including doubtful instances.

⁴⁰ Cf. to frefranne, u getimbre, bediegled, etc., in P. J. Cosijn, Allwestsachsische Grammatik (The Hague, 1888), II, 154–156, 163–164.

⁴¹ Ibid., p. 167. See also Karl Brunner, Altenglische Grammatik, based on Eduard Sievers, Angels. Gram. (Halle, 1951), p. 352.

⁴² Brunner-Sievers, p. 115. Luick, Hist. Gram., pp. 117–118, 266, dates the loss of w in words like drea not later than the 4th century.

⁴³ Sweet, OET, p. 178.

⁴⁴ Occasional examples of all 4 features will be found in Farman's 10th-century Merc, gloss; see E. M. Brown, The Language of the Rushworth Gloss, ii (Gôttingen, 1892), 13,23, 26, 35–36.

⁴⁶ Sweet, OET, pp. 445–447, 449–450; Walter de Gray Birch, Cartularium Saxonicum (London, 1885–93), i, 560563, 583–584.

⁴⁶ The group should be dated ca. 850; see A. Campbell, “An Old English Will,” JEGP, xxxvii (April 1938), 134.

⁴⁷ Cotton Vespasian A. 22; R. Morris, Old English Homilies, EETS, O.S., xxxiv (London. 1867), 217–245.

⁴⁸ Morris, Old English Homilies of the Twelfth Century, EETS, O.S., Liii (London, 1873).

⁴⁹ I omit 2 points (w as abbrev. for wyn- and probable word division of the Latin text of the Urform) as having no direct bearing upon the age of the Urform itself.

⁵⁰ W. M. Lindsay, The Corpus Glossary (Cambridge, 1921), p. 189.

⁵¹ E.g., E. M. Brown, Language of the Rushworth Gloss, p. 15; Brunner-Sievers, pp. 60–61.

⁵² F. Holthausen, Altenglisches etymologisches Wörlerbuch (Heidelberg, 1934), pp. 231, 232.

⁵³ The æ-genitives can be paralleled in EME; for a text nearly contemporary with E, see A. O. Belfour, Twelfth-Century Homilies in MS. Bodley 343, EETS, O. S., cxxxvii (London, 1909). In the last third of this text (about 20 printed pages), I find larpeowæs (106/2), seoreæs (112/25), wuldræs (114/13), heofenæs (124/30), unδancæs (128/12), rænæs (132/7), plus a few ambiguous examples which may have been intended as plurals. There are also æs-genitives in the 10th-century gloss of the Lindisfarne Gospels; e.g., anæs (unius) Matthew pref., p. 14, line 3 (fol. 18^v); domæs (iudicii) ibid., p. 17, line 11 (fol. 20^r); heofnæs (caeli) Mark 4, 32 (fol. 102^v); wercæs (facli) ibid. 5, 14 (fol. 104^r); blindæs (caeci) ibid. 8, 23 (fol. 111^r); deadæs (mortis) Luke 23, 22 (fol. 198^r). For further examples, see A. S. Cook, A Glossary of the Old Northumbrian Gospels (Halle, 1894). References by page and line or by chapter and verse are to W. W. Skeat, The Holy Gospels in Anglo-Saxon, Northumbrian, and Old Mercian Versions (Cambridge, 1871–87). Folio references are to T. D. Kendrick et al., Codex Lindisfarnensis, Vol. i (Basle & Olten, Switzerland, 1956), in which all of the above forms have been verified.

⁵⁴ Cosijn, i, 38–42, 48–65.

⁵⁵ It may, however, have contained elements ultimately traceable to A. Wildhagen believed A to be of the late 9th century. I subscribe to the view, held by some eminent linguists, that A was written early in the 9th century. My reasons for accepting this view are set forth at some length in “The Vespasian Psalter and the Old English Charter Hands,” Speculum, xviii (Oct. 1943), 458–483, and reinforced by the close linguistic and palaeographical resemblance of A to the Lorica Prayer, written by or for Bishop /ESelwald of Lichfield (A.D. 818–830); see Speculum, xxiii, 620, 625–627.

⁵⁶ Biblical Quotations in Old English Prose Writers (London, 1898), p. xxvi; see also p. xxxii.

⁵⁷ Kritische Entstehungsgeschichte des ags. Interlinear-Psalters, in Palaestra, cli (Leipzig, 1926), 5.

⁵⁸ Gneuss discusses this matter (pp. 158–160), rejecting much of Heinzel's genealogical scheme.

⁵⁹ Heinzel (p. 106) called gefrea a poetic word, but I find forms of this verb (i.e., gefreogari) in several prose texts.

⁶⁰ A similar pattern can be observed for most of the items dealt with in the 2 preceding paragraphs.

⁶¹ The later C has 47 gefreogan, 10 alysan, 1 gloss containing both; 23 geswencydnys, 10 geswinc, 3 gedrefydnys; 14 geedleanian, 3 sellan, 1 gyldan, 1 agyldan, 1 edleanian, 1 ge-leanian.

⁶² Lindelöf, Es. Junius 27, pp. 59, 63.

⁶³ PMLA,-LIV, 10.

⁶⁴ Speculum, xxiii, 606.

⁶⁵ Wildhagen, Psalter des Ead., pp. 129–132.

⁶⁶ This tendency of the glossator was noted by Sweet, OET, p. 187.

⁶⁷ Early phonetic loss of final inflexional -n is also a possibility, but a remote one in the dialect of B.

⁶⁸ The letters u and a in this script have distinctly different shapes, and are seldom confused in the gloss. In this respect, they are very different from such pairs as d and , ii and r, which are frequently interchanged.

⁶⁹ Their silence might be interpreted as complete rejection of the copying theory, but arguments from silence are apt to be dangerous.

⁷⁰ They may list a very few items in the gloss (taken out of their linguistic and palaeographical context), with the suggestion that these are archaic and represent the original. But there is no agreement as to which features were copied; e.g., Wildhagen (above, p. 161) mentions o for oe but says nothing of v for u, while Sisam (above, p. 164) cites v for u but disregards the features listed by Wildhagen.