¹ Judith's parentage is uncertain. The Life of King Edward the Confessor, ed. Barlow, F., 2nd ed. (Oxford, 1992), pp. 38–9Google Scholar describes her as the sister of Baldwin IV of Flanders, though John of Worcester and Orderic Vitalis both describe her as his daughter: see the still useful survey of most of the evidence in Freeman, E. A., The History of the Norman Conquest of England, its Causes and Results, 5 vols. (Oxford, 1867–1879) III, 656–8Google Scholar, and the succinct note in Svoboda, R. A., ‘The Illustrations of the Life of St Omer (Saint-Omer, Bibliothèque Municipale 698’ (unpubl. Ph.D. dissertation, Univ. of Minnesota, 1983), pp. 115–16, n. 275.Google Scholar William of Jumièges (Gesta Normannorum Ducum, ed. Marx, J. (Paris, 1914), p. 88)Google Scholar says that Baldwin IV married Eleanor, daughter of Duke Richard II of Normandy, and Judith could have been the child of this marriage and therefore Baldwin V's half-sister. Baldwin IV's first wife, Otgiva, died in 1031. The date of Judith's birth is unknown. She married Tostig by 1051 and apparently bore him children (Life of King Edward, ed. Barlow, , pp. 82–3)Google Scholar, and she was able to bear Welf IV two sons after her marriage in 1070 (or 1071). This makes less likely Decker-Hauff's suggestion that she was born before 1028 as the daughter of Richard III of Normandy (who died in 1027) and Adele of France (who married Baldwin V of Flanders in 1028). See Decker-Hauff, H., ‘Zur älteren Geschichte der Welfen’, Festschrift zur 900-Jahr-Feier des Klosters, 1056–1956, ed. Spahr, G. (Weingarten, 1956), pp. 31–48.Google Scholar Decker-Hauff's hypothesis ignores (as Svoboda pointed out) the early evidence of the Life of King Edwardand takes too literally German descriptions of Judith as queen of England. These erroneous descriptions could as easily have been caused by Judith being the sister-in-law of the future King Harold as by her supposed kinship to William the Conqueror (as Decker-Hauff argues) through her supposed father Richard III.

² The celebrations of Tostig and Judith's wedding in Flanders (in or before 1051) are linked to the exile (after 8 September 1051) of Tostig's father, Godwine, to that country in the Life of King Edward, ed. Barlow, , pp. 82–3. The Godwine family fled England in 1051 after an unsuccessful confrontation with King Edward the Confessor, though they were to return in triumph the following year.Google Scholar

³ The year of Tostig's appointment to Northumbria is given by the Anglo-Saxon Chronicle 1055 DE: Two of the Saxon Chronicles Parallel, ed. Plummer, C. (Oxford, 1892–1899) I, 186 (text)Google Scholar; The Anglo-Saxon Chronicle: a Revised Translation, ed. Whitelock, D., with Douglas, D. C. and Tucker, S. I. (London, 1961, rev. 1965), p. 130 (translation).Google Scholar The Historia Dunelmensis Ecclesie, in Symeonis Monachi Opera Omnia, ed. Arnold, T., 2 vols., RS (London, 1882–1885) I, 94–5Google Scholar, relates the story of Judith and Tostig's devotion and generosity to the see of St Cuthbert, of the attempt by one of Judith's female servants to enter the church of St Cuthbert, and of Judith's subsequent reparation and gifts to Durham, including a crucifix ‘of wonderful workmanship’. The twelfth-century vita of St Oswine (Miscellanea Biographica, ed. Raine, J., Surtees Soc. 8 (Oxford, 1838), 14–15 and 18–20)Google Scholar records Bishop Æthelwine's gift to Judith in 1065 of some of the hair of St Oswine. It has been stated that the names of Tostig and Judith both appear in the Durham Liber uitae (London, BL Cotton Domitian A. vii). The names of Godwine and Tostig are distinguished by being entered in an eleventh-century hand in conspicuous gold capitals on 12v, and the name Judith appears twice (on 43v and 44v), though it cannot be certain that the countess of Flanders was being commemorated in either case. For these, see the transcript in Liber Vitae Dunelmensis, ed. Stevenson, J., Surtees Soc. 13 (Oxford, 1841)Google Scholar and the facsimile (where the Godwine-Tostig entries are clearer than in the manuscript), Liber Vitae Dunelmensis, Surtees Soc. 136 (Oxford, 1923). On 43v, a ‘Juthitta’ is between ‘Walðf, Simund’ and ‘David’ and on 44v a ‘Jutet’ is between ‘Betriz’ and ‘Gode’.Google Scholar

⁴ The Life of King Edward, ed. Barlow, , pp. 52–3Google Scholar and ASC 1061 D (Two Chronicles, ed. Plummer, I, 189 (text)Google Scholar; Anglo-Saxon Chronicle, ed. Whitelock, et al. , p. 73Google Scholar (translation)) speak of the pilgrims being accompanied by Bishop Ealdred. The most recent editor of the life assigned its composition to 1065–7, but for the view that the only safe terminus ante quem is the death of Edith in 1075, see the review by Darlington, R. R. in EHR 79 (1964), 147–8. However opaque and vague the life might be, its author seemed to know the Godwine family.CrossRefGoogle Scholar

⁵ ASC 1065 C (Two Chronicles, ed. Plummer, I, 192Google Scholar; Anglo-Saxon Chronicle, ed. Whitelock, et al. , p. 138).Google Scholar Both the chronicle and the Life of King Edward, ed. Barlow, , pp. 82–3, speak of their retiring to Saint-Omer.Google Scholar

⁶ The date of Judith's marriage to Welf IV is based on inference. Welf cast aside his second wife, Ethelinde, daughter of the outlawed Otto of Bavaria. This could have happened shortly after Otto's fall in 1070. Decker-Hauff has made the plausible guess that the wedding of Welf and Judith took place at the Goslar Christmas assembly in 1070 (‘Zur älteren Geschichte der Welfen’, p. 41 and nn. 47 and 66) when Welf IV was made duke of Bavaria. The late twelfth-century Weingarten addition to the De inventione et translation sanguinis Domini, MGH, SS 15.2 (Hannover, 1887), 923, says that the marriage was arranged through the offices of a kinsman of Welf, the archbishop of Trier. This archbishop (Uoto) died in 1078.Google Scholar

⁷ See C. Buhl, ‘Weingarten-Altdorf’, and Engelmann, U., ‘Zur frühen Verfassungsgeschichte der Abtei’, Festschrift zir 900-Jahr-Feier des Klosters, pp. 12–30 and 49–57.Google Scholar

⁸ Printed in Wirtembergisches Urkundenbuch, 11 vols. (Stuttgart, 1849–1913) I, 302–3 (no. ccxlv).Google Scholar

⁹ The Annalist Saxo (MGH, SS 6 (Hannover, 1844), 694–5) under the year 1066 speaks of her bringing ‘maximam pecuniam et vix credibilem de Anglia secum adduxit, quia interempto Haroldo cuius pecunia Ethwardi regis et ipsius Haroldi, sicut fama est, ad earn devenit’.

¹⁰ The almost contemporary Bernold in his chronicle says she died on 5 March 1094: Bernold, , Chronicon, MGH, SS 5 (Hannover, 1844), 457.Google Scholar The Weingarten Necrology (MGH Necrologia Germaniae, 2 vols. (Berlin, 1888) I, 221–2) says she died on 5 March (the year 1094 in the manuscript is a later addition). The day is a week before the donation (12 March 1094), which is entered in the Fulda gospels. See the recent discussion by Rudolf, H. U., ‘1090 oder 1094 – Wann erfolgte die übergabe der Heilig-Blut-Reliquie’, Festschrift zum Heilig-Blut-Jubiläum am 12. Marz 1094. 900]jahre Heilig-Blut-Verehrungim Weingarten 1094–1994, ed. Kruse, N. and Rudolf, H. U., 2 vols. (Sigmaringen, 1994) I, 24–5.Google Scholar

¹¹ The inventory made in 1753 was first published by Lindner, P., Fünf Professbücher süddeutscher Benediktiner-Abteien. II: Weingarten (Kempten and Munich, 1909), pp. 117–20Google Scholar, and discussed by Löffler, K., Die Handschriften des Klosters Weingarten, Beihefte zum Zentralblatt für Bibliothekswissenschaft 41 (Leipzig, 1912), 147–9. Löffler was the first to try to identify the Treasury books with present surviving manuscripts. The pink skin used to strengthen the binding of A (as of D and other Weingarten manuscripts), possibly at the time of Abbot Berthold, could take A's association with Weingarten back to the thirteenth century.Google Scholar

¹² For the history of the manuscripts before their sale to Thomas Coke, see Löffler, , Handschriften des Klosters Weingarten, pp. 19–26Google Scholar and Irtenkauf, W., ‘Fulda und Weingarten. Zür Säkularisierung der Weingartner Klosterbibliothek’, Von der Klosterbibliothek zur Landesbibliothek. Beiträge zum zweihundertjahrigen Bestehen der Hessischen Landesbibliothek zu Fulda, ed. Brall, A. (Stuttgart, 1978), pp. 339–60.Google Scholar It is clear from the Fulda librarian Boehmapos;s account (printed in Aus Fuldas Geistesleben. Festschrift zum 150 jährigen Jubiläum der Landesbibliothek Fulda, ed. Thiele, J. (Fulda, 1928), pp. 59–88, at 80–4)Google Scholar on which both Löffler and Irtenkauf are based, that Niboyet took the two Morgan books together with the two other Weingarten books which are also now in New York (the Berthold Missal and the Missal of Heinricus sacrista) from Boehm. Niboyet appears to have suggested that they had been handed over to the new Fulda commander General Baron Thiébault, but in the face of the latter's denials it is difficult to establish who stole them (ibid. pp. 80–1 and 84). Holkham Hall records, on some of which the account in Dorez, L., Les Manuscrits à peintures de la Bibliothèque de Lord Leicester à Holkham Hall, Norfolk (Paris, 1908), pp. 5–7Google Scholar is based, indicate that the four manuscripts were offered to Coke via a Parisian dealer, Delahante, in 1818. Patrick McGurk is grateful to the Honorary Librarian of Holkham Hall, Mr D. Mortlock, for further information taken from letters found in vol. VIII of the Holkham Hall copy of the unpublished catalogue of Holkham Hall manuscripts. William Voelkle kindly informed Patrick McGurk that the Holkham Hall visitors' book records Pierpont Morgan's stay there on the night of 10 August 1926.

¹³ The earliest additions are possibly the names added in thirteenth- and fourteenth-century hands at the bottom of p. 238 and the fourteenth-century addition on p. 240.

¹⁴ Harrsen, M., ‘The Countess Judith of Flanders and the Library of Weingarten Abbey’, Papers of the Bibliographical Soc. of America 24 (1930), 1–13, with pl. 1 and figs. 1–2CrossRefGoogle Scholar. One of the supposedly Flemish gospelbooks associated with Judith by Harrsen, Fulda, Hessische Landesbibliothek, Aa. 44, is now accepted as a Fulda product of the mid-eleventh century: see Die Schausammlung der Hessischen Landesbibliothek Fulda. Ein Begleitbuch (Fulda, 1978), p. 18.Google Scholar

¹⁵ Bishop, T. A.M., English Caroline Minuscule (Oxford, 1971), pp. xvi–xvii.Google ScholarMcGurk, P., ‘Text’, in The York Gospels, ed. Barker, N. (London, 1986), pp. 43–63, at 46–7, n. 19. This scribe is reproduced here in pls. IV and VI–X.Google Scholar

¹⁶ See discussion below, pp. 276–7.

¹⁷ See Koehler, W. [and F. Mütherich from vol. IV], Die karolingischen Miniaturen, 5 vols. (Berlin, 1930–1982)Google Scholar, V: Die Hofschule Karls des Kahlen, and McKitterick, R., ‘The Palace School of Charles the Bald’, Charles the Bald. Court and Kingdom, ed. Gibson, M. T. and Nelson, J. L., 2nd ed. (London, 1990), pp. 326–39.Google Scholar

¹⁸ Heslop, T. A., ‘The Production of De Luxe Manuscripts and the Patronage of King Cnut and Emma’, ASE 19 (1990), 151–96.Google Scholar

¹⁹ An example taken at random is the so-called Bohun group of manuscripts. See The Age of Chivalry. Art in Plantagenet England, ed. Alexander, J. and Binski, P. (London, 1987), items 686–91.Google Scholar

²⁰ See below, pp. 275–6.

²¹ See McGurk, , ‘Text’, p. 46, n. 19.Google Scholar

²² See below, pp. 268–73 and 277–80.

²³ The table in McGurk, ‘Text’, p. 46 shows the contents of later Anglo-Saxon gospelbooks, where the four Judith books stand alone among this relatively small sample. As far as is known, contemporary or near-contemporary continental books do not restrict their accessory texts in the same way. Collation of the books shows that the omission was deliberate. In D two pages were left blank after the prefaces for Mark (25v–26r), for Luke (41 v–42r) and for John (65v–66r), but these could not have been enough to hold the omitted chapter-lists for these gospels. Thus the second scribe in D (who wrote all three prefaces) conformed to the restrictive plan for accessory texts of the other Judith gospels. His practice of opening a new quire with the gospel prologue, and of leaving the verso of the evangelist portrait page free of writing probably accounted for the pages which were left blank.

²⁴ The marginal Eusebian numbers added on pp. 5 and 6 in C seem to be a later addition.

²⁵ The interlineated sigla are ‘t’ for the words of Christ, ‘s’ for the discourse of others, and ‘c’ for the narrative. These are found in all the Passion texts with the exception of Luke and John in D. The Passion text is distinguished further in the following places: A, in Mark by a line left blank before ch. XIV with a line of rustics, and in Luke by another blank line and some rustics; D, in Matthew and Luke by a line of rustics, and in Mark by a line of uncials followed by a line of rustics; C, in Mark by a heading in gold rustics ‘passio Domini nostri lesu Christ isecundum Marcum’, a line of rustics, and a cross before XV.42, and in Luke by a line of rustics; and F, John XIX.39 by a cross.

²⁶ McGurk, , ‘Text’, p. 52.Google Scholar

²⁷ Fischer, B., Die lateinischen Evangelien bis zum 10. Jahrhundert, 4 vols., Aus der Geschichte der lateinischen Bibel 13, 15, 17 and 18 (Freiburg, 1988–1991).Google Scholar

²⁸ Fischer, B., ‘Der Text der Quedlinburger Evangeliars’, Das Samuhel-Evangeliar aus dem Quedlinburger Dom (Passau, 1991), pp. 35–42. Here agreements of 99 per cent are usual and are not taken to suggest a common exemplar, though it is true that this fundamental paper was concerned with grouping texts and not with the question of single or multiple exemplars.Google Scholar

²⁹ McGurk, , ‘Text’, pp. 46 and 55.Google Scholar

³⁰ Hanover's Luke and John chapter-lists belong to the Pⁱ family, and London, BL Add. 34890's to the A family. For the study in ibid., selected readings were collated from the two manuscripts, though only those from the London manuscript were published on pp. 56–63.

³¹ Frere, W. H., Studies in Early Roman Liturgy. II: the Roman Lectionary, Alcuin Club Collections 30 (London, 1934), 157–64.Google Scholar

³² Glunz, H. H., History of the Vulgate in England from Alcuin to Roger Bacon (Cambridge, 1933), pp. 140–7Google Scholar. The other four manuscripts were Eadwig Basan's already-mentioned gospels (BL Add. 34890); London, BL Harley 76; Cambridge, Pembroke College 301; and Cambridge, St John's College 73, which is, unlike the others, a post-Conquest manuscript.

³³ Frere, , Roman Lectionary, p. 158.Google Scholar

³⁴ It should be noted that while scribe B wrote all of Trinity, he may not have written the capitulary in the other two. In the lections in the capitulary, Royal and Loan frequently give the wrong evangelist or Eusebian reference, often taking one or the other incorrectly through eyeskip from the lection immediately above. On one occasion Loan compresses two lections by beginning with one and ending with the second, and Royal does this at least twice. The differences in the gospel extracts could be explained by independent and sometimes unconscious alterations of a common exemplar, these differences usually consisting of the omission of some words at the beginning or at the end of an extract.

³⁵ Royal has been collated fully for a later comparison of scribe B's books with the text of the Judith gospels, but only selected chapters from the Loan and Trinity manuscripts have been collated for this examination of the text of the gospels of scribe B.

³⁶ Cronin, H. S., Codex Purpureus Petropolitanus (Cambridge, 1899), pp. xliii, liii and lxiiiGoogle Scholar; J. Gribomont in. the commentary volume to the facsimile of the Rossano gospels, Codex Purpureus Rossanensis, ed. Cavallo, G., Gribomont, J. and Loerke, W. C., Codices Selecti 86 (Graz, 1987), 175–210, at 178–180, 191, 196–8 and 210.Google Scholar

³⁷ McGurk, , ‘Text’, pp. 56–63.Google Scholar The text used by the second scribe is different, being very close over long passages (e.g. Luke XVII.33–end; John X VII. 1–end) to that of the ninth-century, possibly Breton, London, BL Egerton 609. In other chapters there is some agreement with other members of the ‘Irish’ family to which Egerton 609 belongs, though in John I.1–VI.59 there is no striking agreement with any particular family or known witness. D adds at the end of the Luke genealogy the following sortes, not paralleled in any other witness: ‘(qui fuit Dei) et Filii et Spiritus sancti. Qua cum angelis et archangelis, throni, dominationes, principatus et potestates et xxiiii. senioribus, qui non cessant clamare et dicere una voce dicentes: Sanctus, Sanctus, Sanctus, Dominus Deus Sabaoth a quo originum (!) atque initium omni generi humano, salutariam confitentium bonitatem et misericordiam credentium, quia exaltatum nomen suum in super omne quod est celestium et terrestrium et infernorum ipse regem regnantium recolere, cui virtus auxilium liberare credentium benedictum nominare.’

³⁸ McGurk, , ‘Text’, pp. 51–2Google Scholar, where a distinctive error in the Mark preface (not in Trinity) and an unnecessary reading in John are shown to be shared by this group. In Mark ‘Deum’ is substituted unintelligibly for ‘quoniam’ in the concluding phrase: ‘quoniam qui plantat et qui rigat unum sunt, qui autem incrementum Deus est’. In John the phrase ‘qui etsi post omnes evangelium scripsisse dicitur’ is repeated by another ‘tamen post omnes evangelium scripsit’.

³⁹ Glunz, , History of the Vulgate in England, pp. 141–7.Google Scholar

⁴⁰ Herbert Köllner kindly drew Patrick McGurk's attention to a dislocation in the Luke genealogy shared by both F and A. In these books it is arranged in three columns. The genealogy is clearly meant to be read down each column in turn and not across the page line by line, but even so there are serious breaks in the sequence of progenitors. Thus the correct order of ancestors 8–18, for instance, is Matthathiae, Amos, Naum, Esli, Naggae, Maath, Matthathie, Semei, Iosech, Ioda, and Iohanna. In F and A the order is Matthathiae, Naum, Naggae, Matthathie, losech and lohanna, where every other ancestor is skipped, presumably because the scribe of the archetype or exemplar of both was reading down, and not across, a genealogy arranged in two columns.

⁴¹ Some examples should suffice: Matt. II.13 perdendum meum for perdendum eum; II.16 ab ime for a bimatu; n Aegypto for in Aegypto; II.23 in vitate for in civitate; VI.16 culum for in cubiculum; XXVI.60 plum for in templum; Mark I.11 vax for vox; II.4 tecum for tectum. This scribe's Latin was as poor as his script.

⁴² It is worth noting that C changes its layout in Luke and John from one where paragraphs correspond to Eusebian sections to one with fewer paragraphs with the Eusebian sections simply represented by gold initials within a solid block of text. This change could also represent increasing scribal haste. In the ‘York Gospels’ (York, Chapter Library Add. 1 (Christ Church, Canterbury, c. 1000)), for instance, the same change in layout certainly coincides with a more rapidly executed script.

⁴³ The Mark initial page in C reads: ‘Initium evangelii lesu Christi filii Dei vivi sicut scriptumest in Isaiam prophetam.’ Those in D and A read: ‘Initium evangelii lesu Christi filii Dei sicut scriptum est in Esaia propheta.’ Professor Frede kindly tells Patrick McGurk that ‘vivi’ is unique at this point, though it is perfectly acceptable as an echo of the many instances of ‘Dei vivi’ elsewhere in the Bible. The accusative of the prophet Isaiah is a known variant.

⁴⁴ In the following variants, St stands for the reading in the Stuttgart edition: Biblia Sacra iuxta Vulgatam Versionem, ed. Weber, R. with Fischer, B., Gribomont, J., Sparks, H. F. D. and Thiele, W. (Stuttgart, 1969).Google Scholar

⁴⁵ In these books as in many others the opening of Passion texts is distinguished in some way: see above, n. 25.

⁴⁶ One exception is in D where verses 23–6 from John XIII have been added later by a twelfth-century German hand. Small corrections were made by the Anglo-Saxon scribes, probably as they were writing.

⁴⁷ The scribes in D are reproduced here in pls. IV–V.

⁴⁸ Biblia Sacra iuxta Vulgatam Versionem, ed. Weber et al.

⁴⁹ The gospel attached to the Stowe Missal is made up of selected excerpts from John, whilst in the Book of Deer the gospel of John is the only complete text, Matthew probably, and Mark and Luke certainly, being deliberately left unfinished, their texts ending respectively at VII.22, V.35 and IV.1. For Deer, see Hughes, K., ‘The Book of Deer’, in her Celtic Britain in the Early Middle Ages (Woodbridge, 1980), pp. 22–37.Google Scholar

⁵⁰ It must be admitted, though, that the gaps in C had not hitherto been reported.

⁵¹ Omitted phrases and passages are relatively few in Fischer's apparatus. In contrast, in D in Luke the second scribe omits eleven phrases or verses (nine through eyeskip) in addition to the longer omissions noted above.

⁵² Jane Rosenthal is responsible for the following hypothesis on C's use. Note that.the already mentioned Book of Deer has the complete text of John but only incomplete versions of the other gospels.

⁵³ The possible particular associations of Judith with Mary Magdalen will be further examined by Jane Rosenthal in her study of the Crucifixion illumination in A.

⁵⁴ It might be noted that Warren, F. E., ‘The Evangelistarium of St Margaret of Scotland’, The Academy 800 (3 September 1887), 151, thought that the ‘St Margaret Gospels’ were intended ‘not for public church use but for private devotion’. Jane Rosenthal hopes to publish a study in the near future on ‘selective’ gospelbooks in later Anglo-Saxon England.Google Scholar

⁵⁵ Bernold, , Chronicon, s.a. 1094, in MGH, SS 5 (Hannover, 1844), 457–8Google Scholar: ‘Iuditha uxor ducis Welfonis Baioariae, iam diu infirmata, et ea castigatione non parum meliorata, diem clausit extremum iiii. Nonas martii, et apud monasterium, quod maritus eius in honorem sancti Martini in proprio allodio construxit, a Gebehardo Constantiensi episcopo honorifice sepelitur; ad quod monasterium maritus eius dux capellam ipsius tradidit, quae in auro et argento et preciosissimis paramentis mille libras pene valuit, idemque monasterium in possessionibus centum pene mansis melioravit; sicque melioratum et de sua potestate emancipatum sancto Petro sub iure tributario contradjdit, ut apostolicae sedi principaliter subiacerat, et sub eius defensione, ut alia libera monasteria, iure perpetuo polleret.’ The manuscript of this chronicle might be an autograph.

⁵⁶ The books may have been intended for some liturgical use but misreadings in the John gospel for Good Friday might suggest that this text at least was not used: in XVIII.3 ACF have ‘arinis’ for ‘armis’ and ACDF ‘faucibus’ for ‘facibus’; in XIX. 12 C has ‘nos es’ for ‘non es’; in XIX.31 F has ‘erant enim’ for ‘erat enim’, XIX.32 ‘cura’ for ‘crura’. In the Luke Passion week reading (XXII.1–XXIII.53) D has an eyeskip at XXIII.39–40. Francis Newton's hypothesis (which will be mentioned below) associates the Monte Cassino book with the volume known to have been given by the Empress Agnes at the dedication of the new basilica of Monte Cassino, and this was recorded as being richly bound. See Leo, Marsicanus, Chronicon, in MGH, SS 7 (Hannover, 1846), 722Google Scholar: ‘evangelium cum tabula fusili de argento, opere anaglifo pulcherrime deaurato’. Even if it is true, it does not mean that the volume was bound before it left Judith's possession. For the evidence that F had a rich metal cover, see above, p. 254 and below, pp. 278 and 290.

⁵⁷ The probably late-twelfth-century addition to the legend of the Holy Blood, De inventions et translation Sanguinis Domini in MGH, SS 15 (Hannover, 1888), 921–3, with Judith at 923Google Scholar. See the recent discussion of the evolution of the legend by Kruse, N., ‘Der Weg des Heiligen Bluts von Mantua nach Altdorf-Weingarten’, and ‘Die historischen Heilig-Blut-Schriften der Weingartner Klostertradition’, Festschrift zum Heilig-Blut-jubiläum am 12 Marz 1994, ed. Kruse, and Rudolf, I, 57–76 and 77–123Google Scholar. Note that Judith is said not to have acquired the relic of Holy Blood until the death of Baldwin V in 1067.

⁵⁸ This hypothesis would require both scribes in D to have been occasionally deliberately negligent.

⁵⁹ Haseloff, A., ‘Aus der Weingartner Klosterbibliothek’, Deutsche Literaturzeitung (1905), pp. 1998–2002.Google Scholar

⁶⁰ This is implied in most of the literature, starting with Haseloff and Löffler, though it is not always made clear that the entry in the Judith donation, if, as seems likely, it refers to books, speaks of four and not three codices. Weingarten sources later than the 1094 donation transcribed in F associate Judith with the gift of ‘plenariis’: thus, the Weingarten necrology (MGH, Necrologia Germaniae I, 221–32) says ‘Judita dux regina Anglic, hic sepulta, dedit preciosissimum thesaurum ecclesie, Sanguinem Domini, cum reliquiis sanctorum, palliis et plenariis’; the De inventione et translatione Sanguinis Domini, MGH, SS 15 (Berlin, 1888), 923, ‘praecipuum et incomparabilem thesaurum in auro et argento… cum capella et ecclesiasticis ornamentis, palliis, scriniis, plenariis …’Google Scholar

⁶¹ Patrick McGurk is grateful to Dr Peter Dinter of the Mittellateinisches Wörterbuch in Munich for generous help and for providing him with references, including those listed below.

⁶² ‘He gave plenarium of the four Gospels, one epistolary and a sacramentary’: Urkundenbuch des Landes ob der Enns, herausgegeben von Venvaltungs-Auschluss des Museums Francisco-Carolinum zu Linz, 8 vols. (Vienna, 1852–1883) I, 49.Google Scholar

⁶³ ‘If the lord abbot is there, the deacon should bring the incense box to him. The abbot should put the incense in the censer, make the blessing asked for by him, and, as well as making the sign of the cross over the plenarium, he should wave incense over it. After he has read the Gospel, the deacon should give the book to the subdeacon’: William of Hirsau, Constitutiones Hirsaugienses seu Gengenbacenses (PL 150, col. 1022).

⁶⁴ ‘Then we worship as though a book full of the Gospels, which we call plenarium’: Gerhoh of Reichersberg, Tractatus, in Gerhohi opera inedita, ed. D. and van der Eynde, O. and Rijmersdael, A., Spicilegium Pontificii Athenaei Antoniana 8, 3 vols. (Rome, 1955–1956) I, 127.Google Scholar

⁶⁵ These German names on p. 238 of C are listed below, p. 295.

⁶⁶ The hypothesis was generously related to Patrick McGurk by Professor Newton, who gave him a copy of a talk he had given on the subject.

⁶⁷ The names at the foot of p. 238 include ‘Henricus, Adelaita, Henricus, Auta’, and they could be those of Agnes's children who were alive at the time of the dedication of the basilica. These names are written in the lower margin of p. 238, which is the penultimate page of John, and where they are followed by other (mainly German) names (see description on p. 295 and pl. X). Newton suggested that these names were transferred to this page from a now lost fly leaf which was used for visitors’ signatures.

⁶⁸ Patrick McGurk owes this information to Dom Faustino Avagliano, the archivist of Monte Cassino.

⁶⁹ Löffler, , Die Handschriften des Klosters Weingartens, pp. 147–8Google Scholar. The earliest reference to Judith's books being in the treasury is in Hess, , Prodromus monumentorum Guelficorum seu catalogus abbatum imperialis monasterii Weingartcnsis (Kempten, 1781), pp. 64–5, where he speaks of Berthold's Missal ‘in sacrario cum antiquioribus a Juditha Guelfi IV uxore oblatis etiamnum servatur’.Google Scholar

⁷⁰ Some discussion in Needham, P., Twelve Centuries of bookbinding 400–1200 (New York and London, 1979), pp. 33 and 35–8Google Scholar; and in Nixon, H. M. and Foot, M. M., The History of Decorated Bookbinding in England (Oxford, 1992), pp. 19–20.Google Scholar

⁷¹ See descriptions on pp. 297 and 300–1. The skin covers the board binding of both, and is tucked over the inside of the lower back covers.

⁷² In conversation with Patrick McGurk on 31 March 1992.

⁷³ For reproduction of the lower cover and of its stamps, see Köllner, H., Die illuminierter Handschriften der Hessischen Landesbibliothek Fulda. I. Handschriften des 6. bis 13. Jahrhunderts, ed. Köllner, H. (Stuttgart, 1976), Bildband, ill. 948 and 963, nos. 41–7Google Scholar. See the description in Hausmann, R., Die theologiscben Handschriften des Hessischen Landesbibliothek Fulda bis zum Jahr 1600, Die Handschriften des Hessischen Landesbibliothek, Fulda I, ed. Broszinski, H. (Wiesbaden, 1992), 59–61.Google Scholar

⁷⁴ Löffler, , Die Handschriften des Klosters Weingartens, pp. 147–8.Google Scholar

⁷⁵ Swarzenski, H., The Berthold Missal. Pierpont Morgan Library MS 710 and the Scriptorium of Weingarten (New York, 1943), pp. 8–9, figs. 1–3 and 8Google Scholar. See also Jakobi-Mirwald, C., ‘Kreuzigung und Kreuzabnahme in den Weingartner Handschriften des 12. und 13. Jahrhunderts’, Festschrift zum Heilig-Blut-Jubiläum am 12. Marz 1094, ed. Kruse, and Rudolf, I, 185–208.Google Scholar

⁷⁶ In D, the corrections and repunctuation are on 59r and are for Luke XIX.2–10. This passage is the lection for Frere's Standard series‘natalis papae’. In A they are on 141 v–148r and cover John XII.8–XVII.26. The missing words added in the margin are on 146v. In F they are on 86v and cover John XX.19–31. This passage is included in the lection in Frere's Standard series ‘pascha vii.’

⁷⁷ A list of the marked passages is found in the description of the two manuscripts; see below, pp. 290 and 300. The following marginal marks in A correspond to the lections shown: at Matt. XIII.24, the opening of Frere's Standard 34 (fifth week after the Theophany, Sunday); at Luke V.12 and V.15, the opening and the end of Frere's Standard 22 (third week after Theophany, feria vi.); at John VIII.31 and VIII.47, the opening and the end of Frere's Standard 288 (‘pro ordinationibus’ or other alternatives). There are very few marks in F, all being made by the stylus, and only Matt. VII.15 (‘ebd. iii. post natal, apostolorum’) and Matt. X.37 (St Laurence) correspond to masses in Frere's Standard series, though some alternative ferias might possibly explain some others. It might be worth noting that in Matt. X.37, the passage marked by a stylus is defective since it almost immediately opens with a phrase omitted through eyeskip: ‘et qui amat filium aut filiam super me non est me dignus’.

⁷⁸ The clearly liturgical marks added in Tironian notes in the ninth century to the Italian sixth-century gospels, London, BL Harley 1775, for instance, sometimes only show the end of a lection. It is likely that the opening would have been easily identified by the marginal Eusebian references.

⁷⁹ For marks for the reading of the Passion, see above, n. 25.

⁸⁰ Gold initials mark the unnumbered Eusebian sections. The brown initials used intermittently by the second scribe in D never announce such sections.

⁸¹ On compass points, see the introduction, by Gullick, M., to A Twelfth-Century Working Alphabet of Initial Letters from Tuscany (Hitchin, 1979), p. 8, nn. 1, 9 and 11.Google Scholar

⁸² ‘Quire-tacketing’ is Michael Gullick's term, and is studied in his forthcoming paper in the Roger Powell memorial volume.

⁸³ This is on the likely assumption that the same person was responsible for the gold capitals in the text marking Eusebian sections and for the display capitals. One hand is responsible for the text capitals in C and F. These are well balanced and orderly, have relatively few serifs, and are notable for a slightly squashed and rounded E and C, and the display capitals in C (but not of course the continental ones in F) share the same features. In A the text capitals are sometimes weak and poorly balanced, and often have a straggling capital A, and these characteristics are shared by A's display capitals. In D the text capitals in both scribes' stints were unlike those in A, C and F, being rather heavy and serifed, with a distinctive A and an I with no serif on the top right or bottom left of the upright, and were probably the work of D's second scribe. D's display capitals are distinctive, and it is just possible that they were also his work. The gold minuscule script on the Matthew and John initial pages in C was certainly written by the scribe of that manuscript but it is doubtful whether the gold writing on the open books held by Matthew and John in A were the work of the poor scribe. For reproductions of the initial pages in ACD, see Anglo-Saxon Textual Illustrations. Photographs of Sixteen Manuscripts with Descriptions and Index, ed. Ohlgren, T. H., Medieval Institute Publ., Western Michigan Univ. (Kalamazoo, MI, 1992)Google Scholar where A, 3r, 49r, 78r and 123r are found in pls. 11.4, 11.6, 11.8 and 11.10; C, pp. 3, 103, 127 and 167 are in pls. 13.2, 13.4, 13.6 and 13.8; and D, 3r, 27r, 43r and 67r are in pls. 12.3, 12.5, 12.7 and 12.9 respectively.

⁸⁴ The identification of the illuminator in part of that sacramentary with that in F can be best approached via Rhin-Meuse. Art et civilisation 800–1400 (Cologne and Brussels, 1972), pp. 229–30 with bibliography. Herbert Köllner first suggested to Patrick McGurk that the first scribe of the Paris sacramentary also wrote in F. This scribe certainly wrote the gold script in F and was almost certainly responsible for the purple background on the first pages of Matthew in F. One scribe-artist was probably at work in both books.Google Scholar

⁸⁵ It should be noted that the second quire in A and F does not begin with the same text. This would not be surprising, since even if the same exemplar was being followed by both books the layout and the number of lines in the two were not identical.

⁸⁶ In A the frontispiece illumination of the Crucifixion is an integral part of the quire where the Matthew portrait is on an inserted leaf.

⁸⁷ For Eadwig Basan, see Bishop, , English Caroline Minuscule, pp. xv and 29 and pl. xxiiGoogle Scholar, and Dumville, D. N., English Caroline Script and Monastic History (Woodbridge, 1993), pp. 111–40Google Scholar. Charters with comparable script include London, BL Add. Ch. 19801 (S 1405), dated to 1058 (Facsimiles of Ancient Charters in the British Museum, ed. Bond, E. A., 4 vols. (London, 1873–1878) IV, 38)Google Scholar, and Hertford, Hertfordshire Record Office, D/ELW Z22/4 (S 1031), dated to 1060 (Anglo-Saxon Charters. Supplementary Volume I. Facsimiles of Anglo-Saxon Charters, ed. Keynes, S. (Oxford, 1991), no. 22, pl. 22).Google Scholar

⁸⁸ Temple, Anglo-Saxon Manuscripts 900–1066, no. 79, pls. 259–60, illustrates both illumination and script. On p. 110 Temple compares the illumination of Douce 296 with that of D.

⁸⁹ Many eleventh-century initial pages in luxury Anglo-Saxon books are unfinished or poorly planned. Examples include: London, BL Add. 34890, fol. 74 (Luke page) with a decorated initial but no text; Cambridge, Pembroke College 301, fol. 71 (Luke page: ibid. pl. 236) with much text missing; London, BL Arundel 155, fol. 53 (Ps. LI: ibid. pl. 217) with text in crowded rustics and lines left blank; Royal 1. D. IX (John page: ibid. pl. 222), where the heavily decorated frame dominates the illumination. The capitals page in the Edgar New Minster charter, London, BL Cotton Vespasian A. viii, 4v, is reproduced in McGurk, P., ‘The Golden Age of Anglo-Saxon Painting 966–1066’, Miniatura 1 (1988), 163–6, at 163Google Scholar. The comparison between the display capitals in this charter and in D was made independently by Gameson, R., ‘English Manuscript Art in the Mid-Eleventh Century: the Decorative Tradition’, AntJ 71 (1991), 64–122, at 77 and 84–5, and figs. 14–15CrossRefGoogle Scholar, which appeared after this paper had been accepted for publication, and in which further interesting observations are made on the Judith books on pp. 77 and 88.

⁹⁰ As well as being earl of Northumbria, Tostig was earl of Nottingham and earl of Northampton. On the evidence for this, see Harmer, F. E., Anglo-Saxon Writs (Manchester, 1952), pp. 418, 262–5 and 574Google Scholar; Williams, A., ‘The King's Nephew: the Family and Career of Ralph, Earl of Hereford’, Studies in Medieval History presented to R. Allen Brown, ed. Harper-Bill, C. et al. (Woodbridge, 1989), pp. 327–43, at 333, n. 33 and 338–9Google Scholar. In a letter dated 15 May 1989, A. Williams kindly summarized for Patrick McGurk her findings for Tostig's Domesday Book holdings. He had substantial lands in the North, particularly in Amounder-ness; in the East Midlands, especially Bedfordshire and Buckinghamshire; in the Isle of Wight and the adjacent mainland; and in his brother Harold's western earldom, Gloucester shire and Oxfordshire. See now Fleming, R., Kings and Lords in Conquest England (Cambridge, 1991), pp. 46, 56, 70–1, 86–91 and 102CrossRefGoogle Scholar, and Clarke, P. A., The English Nobility under Edward the Confessor (Oxford, 1994) pp. 24–5 and 191–4Google Scholar. Ann Williams kindly informed Patrick McGurk of the absence of evidence for Judith's landholdings in Domesday.

⁹¹ The Thorney Liber uitae is London, BL Add. 40000. It is discussed in Gerchow, J., Die Gedenküberlieferung der Angelsachsen mit einer Katalog der libri vitae und Necrologien, Arbeiten zur Frühmittelalterforschung 20 (Berlin and New York, 1988), 187–99Google Scholar. On his flight from England in 1051, Godwine embarked from Thorney (Anglo-Saxon Chronicle D, s.a.). This has sometimes been mistaken for the monastery of Thorney, but it is almost certainly Thorney near Bosham in Sussex where Godwine had many estates. It is also just possible that another Countess Judith, Earl Waltheof's wife, who held many lands near Thorney, may have been mistaken by Harrsen for the subject of this paper. Harrsen (‘Countess Judith of Flanders and the Library of Weingarten Abbey’, p. 4) claims that Tostig had helped the endowment of Thorney abbey. Note that ‘Wltheth comes et uxor eius’ are in the Thorney Liber uitae.

⁹² The models for the Judith evangelist and initial pages will be considered in detail by Jane Rosenthal in a later publication.

⁹³ Judith's collecting zeal is demonstrated by her bequest to Weingarten in which the ‘plenaria cum uno textu evangelii’ were only four among very many more religious objects. For the evidence of professional scribes and craftsmen in the Anglo-Saxon period, see Dodwell, C. R., Anglo-Saxon Art (Manchester, 1982), pp. 68–83Google Scholar, and for professional scribes at slightly later dates, see Gullick, M., ‘The Scribe of the Carilef Bible: a New Look at some Eleventh-Century Durham Manuscripts’, Medieval Book Production: Assessing the Evidence, ed. Brownrigg, L. (Los Altos, CA, 1990), pp. 61–83, at 76 and 83.Google Scholar

⁹⁴ The northern sources are those mentioned above, n. 3, namely the Historia Dunelmensis ecclcsie and the Vita S. Oswini.

⁹⁵ Vita S. Oswini, pp. 19–20.

⁹⁶ Patrick McGurk has been responsible for this first part of a study on the Judith Gospels, though Jane Rosenthal contributed the hypothesis òn the Monte Cassino book's function. A later publication will focus on the illumination and metal covers and will be largely the work of Jane Rosenthal. Throughout there has been a fruitful and extremely useful collaboration between the two. Patrick McGurk has incurred many obligations which he would gratefully like to record: to the British Academy, which made a grant for the first of two visits to Fulda and Monte Cassino; to the librarians of the Staatliche Bibliothek, Bamberg, the Hessische Landesbibliothek, Fulda, the Archivio della Badia, Monte Cassino, the Pierpont Morgan Library, New York, the Bibliothèque Nationale, Paris, and the Württembergische Landesbibliothek, Stuttgart for allowing the repeated examination of precious codices; to the kindness and help of Dom Faustino Avagliano at Monte Cassino, of Eva-Marie Thole at Fulda, of Felix Heinzer at Stuttgart. He would like to thank Michael Gullick for his comments and guidance on quire tacketing, compass points and different ways of preparing and applying gold in manuscripts; Sandy Heslop for a useful discussion on the problems of the Judith gospels; Christopher Clarkson for reporting his observations of the bindings of the two Pierpont Morgan books; and John Mitchell for the photographs of the Monte Cassino codex. He very much regrets that Herbert Köllner did not live to see this article in print, as it would have been a small return for the assistance and advice he always generously and selflessly offered. Both Patrick McGurk and Jane Rosenthal would like to express their appreciation of William Voelkle's willingness to allow them to examine more than once the two valuable and fragile Judith codices which are among the jewels of the Pierpont Morgan Library in New York.