One of the oldest and most important manuscript books surviving from the Cathedral priory of St. Andrew is the Gundulf Bible, now designated MS. HM 62 in the collection of the Huntington Library, San Marino, California, USA. A two-volume great Bible copied c.1075, it purports to have been the gift of Gundulf, the second Norman bishop of Rochester (1077-1108). He is linked to the Bible by a 13th-century inscription on the flyleaf (fol. 1r) of each volume as shown in plates 1 and 2. These inscriptions are significant for their early date and placement at the head of the flyleaf. The usual Rochester ex libris and ex dono are from the 14th century and occur in that order at the foot of the flvleaf.

Thus the attribution to Gundulf seems to represent a living tradition rather than part of a larger cataloguing effort.

A long description of the Bible appears as item 26 in the first medieval catalogue of the Rochester library, entered about A.D. 1124 into the second part of the Textus Roffensis (fol. 225v), the Anglo-Saxon cartulary of the priory of St. Andrew:

Vetus et novum testamentum quam transtulit de hebreo in latinum in i voluminibus. Quorum primum continet hos libros. Quinque libros moysi. lesum nave. ludicum. Ruth. Psalterium. Proverbiorum.

Acclesiastes. Sapientize.

Ecclesiasticum. Hezram et neemiam. Paralipomenon duos libros. et ii evangelia. In alio vero volumine continentur iii libri regium. lob. Liber tobie. ludith. Hester. Libri machabeorum i. Libri prophetarum omnes. Actus apostolorum. Epistolae pauli. aliorumque apostolorum. Apocalypsis.

The same entry occurs in a fragmentary catalogue from the 12th century, now part of the archive of the Dean and Chapter of Rochester Cathedral (DRc/Z18). The latest medieval description appears as item 48 in the catalogue of A.D. 1202, entered on two leaves added to a volume of Augustine's works, British Library MS. Royal 5 B. XII.

Medieval cataloguers did not connect the Bible to Gundulf nor indicate that it had any particular significance. The only external evidence of his possible role in its acquisition occurs on fol. 123r of British Library MS Cotton Vespasian A. XXII, an early 13th-century collection of materials related to the priory, where Gundulf is memorialized for having built Rochester castle and established a library of sixty books. Nevertheless a strong case can be made for local production of the Bible at a time when the newly refounded priory had need of such. Its plain appearance and lack of decoration, along with an unusual arrangement of the contents, suggest that utilitarian motives lay behind the acquisition, whether or not Gundulf was involved. The Bible's most unusual feature, for example, is the separation of the Gospels and Pauline Epistles, each placed at the conclusion of volumes I and II respectively. One explanation for this arrangement was the possible need for the Bible to serve double duty as a Gospel Book and Epistolary in services if books were in short supply, as they appear to have been in the immediate post-Conquest period.

Why then is this Bible so important to the history of Rochester's medieval library? First, it is the earliest Bible recorded in the collection. Second, as will be shown in the course of this paper, the Gundulf Bible embodies a particular type of Vulgate text adopted from Christ Church, Canterbury and used for as long as two centuries at Rochester. Moreover, it sheds light on the other surviving Vulgate and Gospel books from Rochester, showing the persistence of a significant textual tradition localized to Kent.

The size and layout of the Gundulf Bible are typical of great Bibles from Norman England.

Though the margins have been trimmed, the pages still measure 400 × 265mm with a written space of approximately 340 × 200mm ruled by dry point in double columns. it has neither illuminations nor decorated initials, only a few modestly embellished and coloured capitals in red, purple, yellow, and blue. Red display script is used sparingly except in the Psalter, where it dominates. The main text has been written in a Norman script style of the late 11th century. Its most unusual feature is a horned e, often thought to be an early English feature, but persisting beyond the Conquest at Exeter and Winchester, for example. There are numerous marginal corrections and annotations in contemporary and later hands. The volumes are roughly equivalent in length (I has 240 leaves, Il has 262), though the collation indicates that I-IV Kings were copied separately from the rest of volume Il and bound with it possibly to equalize the set. This arrangement had been made by the time of the earliest catalogue description, quoted above.

The Gundulf Bible is linked most directly to Canterbury through its close relationship to pre-and post-Conquest Vulgates associated with Christ Church. These are best understood as a family of texts with many similarities and a few individual features.

Although it seems to have developed from continental books imported during the 10th-century Benedictine Revival, the tradition resulted in one complete two-volume great Bible surviving for Canterbury and for Rochester, one partial decorated set for each foundation, and the majestic Dover Bible (Cambridge MS Corpus Christi College 3-4) set to Christ Church's dependency there'. Some Gospel Books with Canterbury connections share in this textual type, but not so 'Goda's Gospels' (British Library MS Royal 1 D. III), the only survivor from Rochester. An 11th-century book once owned by Godgifu, half-sister to Edward the Confessor, the Gospels did not reach Rochester until the reign of William II (1087-1100), well after the production of the Gundulf Bible.

British Libraray MSS. Royal 1 E. VII-VIII, copied in the late 10th century probably at Canterbury, comprise the two-volume great Bible most closely related to the Gundulf Bible. Both share a large size, two-column format, plain appearance and a textual type that can be traced to Vulgates produced in Northern France during the 9th century and brought by monks to the Benedictine house of St. -Germain-des-Prés near Paris during the Viking invasions of the 9th-10th centuries, where they were studied and reproduced.

Additional similarities include: a peculiar order of prefaces to the Gospels, with identical chapters; an extremely rare incipit to the chapters of Numbers (Numerantur ex precepto); the lack of canon tables; a unique series of prefaces to Romans; a long prologue to 1 Corinthians (Corinthus metropolis civitas Achaiae); Laodicians following Hebrews. There are sufficient differences between MSS. Royal 1 E. VIl-VIlI and the Gundulf Bible to preclude exact copying, for example the chapters to Joshua-Judges and 1-II Kings, and readings of particular verses, but the affinity of the texts is clear. Apparently because of their close relationship to the neighbouring foundation, the monks at Rochester adopted a Vulgate text established at Canterbury, made certain modifications for their own purposes, and maintained it into the 13th century.

Of special interest, therefore, are the innovative features of the Gundulf Bible and later additions made at Rochester in the distinctive 'prickly' script style of the early 12th century. The order of books is apparently unique. No similar order is recorded by Berger nor found in any descriptions of the medieval Vulgate consulted in the course of this study?. Volume I has the Octateuch, Psalms, books of Wisdom, I-II Paralipomenon (Chronicles in the King James version of 1611), Esdras, Nehemiah, and the Gospels.

Volume Il opens with I-IV Kings, followed by Job, Tobias, Esther, Judith, I-II Maccabees, the Prophets, Acts, Catholic Epistles, Apocalypse, Pauline Epistles. This order is verifiably original as indicated by the continuity of the text and the first catalogue description, although Apocalypse is mistakenly recorded there as the last book in Volume I.

Apocalypse cannot have been rearranged subsequent to the catalogue description because the prologues to the Pauline Epistles immediately follow the explicit (fol. 288г).

As suggested above, there must have been practical reasons for the unique organization of the Gundulf Bible. In fact, rearranging texts was a common practice at the Rochester scriptorium dating from the Anglo-Saxon period. Traditional formats became secondary to new purposes as shown, for instance, in revisions to Old English homiletic collections'.

Revision is most evident in plain books, such as the Gundulf Bible, intended for ordinary use.

Twelfth-century additions to the Bible further reflect a lack of concern for customary formats. Dedicatory verses by the two most famous Carolingian revisers of the Vulgate, Alcuin and Theodulf, appear at the end of Volume I. Normally found in 9th-century Bibles from Tours and Orleans respectively, they represent rival versions and never occur together as far as we know. The verses are rare in English Bibles; Alcuin's are found in one 13th-century copy, whereas Theodulf's do not appear in any other known Bible of English origin. In the Gundulf Bible, Alcuin's poem is the reduced version of 18 lines also found in several continental manuscripts. Theodulf's poem has been edited uniquely, the last four lines - a kind of envoy - being omitted. We can only speculate about the purpose for including these verses. They are copied without attribution, and they do not work as indices because the order of books cited in the verses does not reflect the peculiar arrangement of the Gundulf Bible. At most they serve as mnemonic devices for learning the books in any Bible. As a final bizarre touch, the lines omitted from Theodulf have been filled by a later hand with six lines of Anglo-Norman verse, the opening to a lost chanson de Geste. Moving to volume II, additions include the prologue and text to the apocryphal Book of Baruch followed by the letter of Jeremiah in an Old Latin version normally found in early Italian and Spanish Bibles. These materials further indicate the availability of continental sources to Rochester copyists in the early 12th century, from which they drew to augment their Vulgate.

Even as they were adding to the Gundulf Bible, scribes at Rochester produced a typical monastic library of service books, patristic works, histories, homiliaries, and records of their own foundation. The first catalogue lists 98 items, some in multiple volumes.

Although they relied on Canterbury to provide many of the texts they copied, the Rochester contingent developed their own distinctive styles of script and illumination apparent today in the surviving books. Moreover, by 1125 a new multi-volume decorated Vulgate was underway, two volumes of which remain, the New Testament (Baltimore, Walters Art Gallery MS. W. 18) and Joshua-Judges-Ruth-Kings (British Library MS. Roval 1 C. VII).

The text of the new set is very close to that of the Gundulf Bible, and suggests that there was a commitment to preserving the standard adopted from Christ Church, Canterbury.

As regards prefaces and chapters, the Walters New Testament lacks quire 1 and whatever prefaces to the Gospels it contained. Further, no prefaces and chapters are assigned to the individual Gospels. In subsequent books, however, the prefactory materials are identical to those found in MS. Royal 1 E. VIll and the Gundulf Bible, including the lengthy series preceding Romans. With regard to the portion of the Old Testament preserved in MS. Royal 1 C. VII, chapters to Joshua and the preface to Judges reflect those found in the Canterbury Bible, MS. Royal 1 E. VIl, and the preface to Kings found in both earlier Bibles, whereas chapters to Kings are omitted. There are some independent readings of individual verses in this new Vulgate, but the text is very close to that of the Canterbury and Rochester Bibles. Where the earlier Bibles disagree, the new version shares approximately equal numbers of readings with one against the other. In terms of decoration, both remaining parts of the new set draw on a common group of motifs including the human-profile terminal, flowers, fruit, and griffins. MS. Royal 1 C. VIl has, in addition, four coloured historiated initials at the openings of Joshua and I, II, IV Kings. Space for one such initial has been left at the opening to Matthew in the Walters New Testament, but it was never filled. The evidence of the library catalogues and surviving books indicates that this set, originally in five volumes, remained the sole decorated Bible at Rochester for at least a hundred years. Late in the thirteenth-century the priory acquired a Vulgate, now known as Oxford MS. St. John's College 4, which departs from the distinctive text found in the earlier Rochester books.

How did the Rochester books come to be disbursed, some to journey to the United States? Working with booklists from the 16th century providing the foundations of the Royal Library, James P. Carley has connected the great number of manuscripts from Rochester Cathedral Library appearing on these lists with the seizure of St. John Fisher's books in 1534. Somehow the Gundulf Bible came into the hands of Lord Lumley, for it is described in a catalogue of his collection made c. 1611 as Biblia vetusta quondam Gundolphi episcopi Roffensis*.

It seems not to have been absorbed into the Royal Library, as were many of Lumley's books. In succeeding centuries it was owned privately first in the low countries and later in England. According to C. W. Dutschke, who has catalogued the medieval manuscripts in the Huntington Library, the Bible belonged for a time to the Amsterdam theologian Herman van de Wall (1672-1733). After his death the library was sold in 1734, and the Bible changed hands at least four times before coming to the Rev. Theodore Williams, who had it rebound. In 1827 Sir Thomas Phillipps purchased the Bible in London from the Williams collection, Phillips n. 3504. Then in 1924 Mr. Thomas Fitzroy Fenwick, grandson of Sir Thomas, sold the Bible and other materials to the bookseller Dr. A. S. W. Rosenbach who, in turn, resold many of them to the late Henry L. Huntington, in whose library the Gundulf Bible resides todayS.

The two surviving volumes from the five-part set have divergent histories. Henry Walters from Baltimore purchased the New Testament early in the twentieth century in Paris from Leon Gruel. Loss of family records makes the exact date unrecoverable. On the other hand, MS. Royal 1 C. VII bears the old Royal press-mark number 507, which corresponds to the matching description in the Westminster Palace inventory of 1542. This volume must have been taken directly into the Royal Library.

Entries in Rochester's medieval catalogues indicate only the two complete Vulgates described in this paper as being part of the library, though reference is made to Psalters, Pentateuchs, and glossed versions of individual books. More startling is the lack of Gospel books, none mentioned in any inventory of the collection. Given the great number of surviving Gospels associated with Canterbury and the omission of reference even to Goda's Gospels in the Rochester catalogues, a book verifiably in possession of the priory, we must assume that these were stored apart from the items inventoried, perhaps in the Cathedral itself, and were not considered part of the library. It would be interesting to know how the Gundulf Bible escaped the Royal Library as perhaps did the Walters new Testament. Our best hope is that the work of James Carley and others on the early history of the Royal collections will help in tracing their journeys.

Mary P. Richards
University of Delaware

Notes

Cambridge MS. Pembroke College 301, c. 1020, attributed to Canterbury by Temple (#73, pp. 91-92) preserves a portion of the New Testament (Gospels and preface to Acts) with strong textual links to MS. Royal 1 E. VII-VIII. This is the part of the decorated Vulgate surviving from Christ Church which parallels the Walters New Testament for Rochester. Although unfinished, the decoration of MS. Pembroke College 301 is far more ambitious than that of the later Rochester set, for it includes elaborate canon tables and full page Evangelist portraits.

On the basis of the type and order of prefatory materials and individual readings provided by Glunz, the Vulgate text localized to Kent can be distinguished from those at Lincoln, Durham, Bury St. Edmunds, St. Albans, and Winchester. Glunz, however, is mistaken in the belief that the Gundulf Bible comprises a revision made by Lanfranc.

See Mary P. Richards, 'Innovations in Alfrician Homiletic Manuscripts at Rochester', Annuale Mediaevale 19 (1979): 14-26.(4) Printed in the Concordance with Lord Lumley's Catalogue, Catalogue of Western Manuscripts in the Old Royal and King's Collection, ed. George F. Warner and Julius P. Gilson (London, 1921), 1:X1. See also item 111, p. 48, in The Lumley Library: The Catalogue of 1609, ed, Sears Jayne and Francis R. Johnson (London, 1956).

Information provided by Mr. D. A. H. Cleggett from A. N. L. Munby's Phillipps Studies, 5: 78-79.

The early 13th-century donation list for Rochester (British Library MS. Cotton Vespasian A. XXII, fols. 88г-v) mentions that Goda's Gospels had been redeemed from mortage by Prior Helyas about 1200.

Bibliography

Berger, Samuel. Histoire de la Vulgate. Paris, 1893.

Carley, James P. 'John Leland and the Foundations of the Royal Library: The Westminster Inventory of 1542', Bulletin of the Society for Renaissance Studies 7.1 (1989): 13-22.

Coates, R. P. 'Catalogue of the Library of the Priory of St. Andrew, Rochester, from the Textus Roffensis', Archaeologia Cantiana 6 (1864-65): 120-28.

Dutschke, C. W. Guide to Medieval and Renaissance Manuscripts in the Huntington Library (San Marino, CA, 1989), 1: 124-30.

Glunz, Hans. History of the Vulgate in England from Alcuin to Roger Bacon. Cambridge, 1933.

Ker, N. R. Medieval Libraries of Great Britain. 2nd ed. London, 1964.

Richards, Mary P. 'A Decorated Vulgate Set from 12th-Century Rochester, England', The Journal of the Walters Art Gallery 39 (1981): 59-67.

Texts and Their Traditions in the Medieval Library of Rochester Cathedral Priory.

Transactions of the American Philosophical Society 78:3. Philadelphia, 1988.

Rye, W. B. 'Catalogue of the Library of the Priory of St. Andrew, Rochester, A.D. 1202* Archaeologia Cantiana 3 (1861): 8-15.