Codex Usserianus Secundus (MS 56)

Introduction and Scope

Among the seven early medieval Gospel books preserved in the Library of Trinity College Dublin, the manuscript catalogued as MS 56 (A.4.6) and known variously as the Codex Usserianus Secundus, the Garland of Howth, and the Ceathair-Leabhair occupies a peculiar and instructive position. It is at once one of the most textually significant and one of the least art-historically studied of the group. Overshadowed by the Book of Kells and the Book of Durrow, and long treated as a poor relation even of its near-namesake the Codex Usserianus Primus, the Garland of Howth has attracted sustained attention chiefly from biblical textual critics, who prize it as a witness to the Old Latin (Vetus Latina) tradition of the Gospels, and only recently from the codicologists and art historians who conserved and digitised it. This essay, conceived as a companion to an earlier dossier on the Codex Usserianus Primus, attempts a synthetic account of what is securely known, what is scholarly consensus, what remains debated, and what is mere speculation or repeated error concerning this manuscript.

The methodological posture adopted throughout is deliberately cautious. Every assertion is layered epistemically: the reader is told, so far as the sources allow, whether a claim rests on the physical object itself, on the near-unanimous judgement of specialists, on live scholarly disagreement, or on tradition and conjecture. Where popular and even semi-scholarly accounts have propagated errors — confusions with the Codex Usserianus Primus, confusions with the entirely unrelated Book of Howth, misattributed dates, and over-confident statements about the manuscript’s origin — these are identified and corrected rather than silently reproduced. Because the requester has stipulated that no more than five named scholars appear in the finished essay, the discussion foregrounds a small number of figures whose work is genuinely load-bearing: T. K. Abbott and H. C. Hoskier, the manuscript’s principal editors; E. A. Lowe, whose Codices Latini Antiquiores fixed its palaeographical place; H. A. G. Houghton, its most recent textual analyst; and Rachel Moss, who led the art-historical reassessment. Other scholars whose contributions are relevant are referred to by their role rather than by name.

Names and Nomenclature: Untangling a Tangle of Titles

Few manuscripts of comparable importance carry so many names, and the multiplicity has itself been a source of confusion. The volume is most romantically called the Garland of Howth. This English title is not, as it might appear, a reference to any physical garland or to the decorative luxuriance of the book; it is a corruption. The consensus of scholarship holds that “Garland” is an anglicisation of the Irish phrase Ceithre Leabhair (variously Ceathair-Leabhair), meaning “four books” — that is, the four Gospels — a common Irish way of designating a Gospel book. The phonetic slippage from the Irish, pronounced roughly “keire lower,” into an English “garland” is the sort of folk-etymological reshaping that frequently attends objects passing between linguistic communities. The association with Howth, the peninsula and village on the northern arm of Dublin Bay, reflects the manuscript’s later medieval and early modern home rather than necessarily its place of origin.

The learned Latin name Codex Usserianus Secundus — “the second book of Ussher” — derives from the seventeenth-century Archbishop James Ussher (1581–1656), the great Church of Ireland scholar, biblical chronologist, and collector whose library ultimately passed to Trinity College. The “Secundus” distinguishes it from the Codex Usserianus Primus (MS 55), the older and textually purer Old Latin Gospel book with which it has always been paired in the collection. It should be stated plainly, because the point is frequently muddled: the two Usserian codices are distinct manuscripts of markedly different date, appearance, and textual character, united chiefly by their shelving history and by the Ussher name.

A recurring and more serious confusion — one this essay actively debunks — is that between the Garland of Howth and the so-called Book of Howth. The two have nothing to do with one another. The Book of Howth is a sixteenth-century compilation of historical and literary material, part of the Carew manuscripts preserved in Lambeth Palace Library, associated with the St Lawrence family of Howth and calendared among the Carew Papers by nineteenth-century editors. It is a late secular codex of Anglo-Irish history; the Garland of Howth is an early medieval Latin Gospel book. Their shared toponym and their common connection to the Howth locality have led careless writers to conflate them, but no bibliographical relationship exists.

Finally, in the technical apparatus of biblical textual criticism the manuscript carries sigla of its own. In the older system inherited from the Latin New Testament scholars it is designated r2 (with the Codex Usserianus Primus as r or r1); in the Beuron numbering maintained by the Vetus Latina Institute it is Vetus Latina 28 (VL 28). In the register of Latin Gospel books compiled in the mid-twentieth century it appears as number 85, and in Lowe’s Codices Latini Antiquiores it is CLA II.272. These sigla matter because they are the coordinates by which the manuscript is located in the scholarly literature, and because a further common error — the misassignment of the Primus and Secundus sigla — has occasionally crept into reference works. To be precise: the Codex Usserianus Primus is r1 / VL 14 / CLA II.271; the Garland of Howth is r2 / VL 28 / CLA II.272.

The Naming History in Detail: From “Kerlower” to Ussher

The documentary history of the manuscript’s name repays closer attention, because it also constitutes part of the evidence for its provenance. According to the tradition reported in the scholarly and antiquarian literature, an early sixteenth-century ecclesiastical record — a visitation of the church on Ireland’s Eye by Archbishop John Alen of Dublin — refers to the book under a form such as “Kerlower,” itself evidently a phonetic rendering of Ceithre Leabhair. This detail, if accurate, is valuable: it indicates that the book was already known by its Irish “four books” designation, and that it was physically present in the Howth locality, by the early 1500s. It was subsequently, in the seventeenth century, that Ussher’s association gave rise to the alternative learned name. The claim that Ussher personally examined the book at Howth and that the English “Garland” title crystallised around this period is repeated in tertiary accounts; the reader should treat the precise sequence with some caution, since the tertiary sources that state it most confidently (for example, the AI-generated compendium Grokipedia) do not cite a controllable primary reference, whereas the underlying facts — the Alen visitation record and the Ussher connection — are grounded in older antiquarian scholarship.

It is worth registering here a parallel with the Codex Usserianus Primus, whose name likewise honours Ussher. In the case of the Primus, it has been observed that despite the “Ussher” title it is actually doubtful whether Ussher ever owned the manuscript in a strict sense; the name was fixed by later cataloguers, above all by Abbott in 1884. A similar caution applies to any assumption that the Secundus was in a strong sense Ussher’s personal property. The Ussher name is a cataloguing convenience and a nod to the archbishop’s central place in the formation of the Trinity collections, not necessarily a statement of documented ownership.

Codicological Description: Support, Dimensions, Layout, and Script

The physical description of the manuscript can be given with a good degree of security, since it rests on direct examination recorded in the standard catalogues and, most recently, on the conservation campaign and digitisation of the mid-2010s.

The Garland of Howth is written on parchment (vellum), the prepared animal skin standard for Insular book production. In its present, much-reduced state it consists of 86 folios. The leaves measure approximately 241 × 192 mm — the figure most often cited, rounded, is about 24.3 by 17.5 cm — with a written area, or text block, of roughly 20 × 15 cm. This is a substantial but not grand format: considerably smaller than the great display Gospels such as Kells, but far larger than the Irish “pocket gospels” such as the Books of Dimma and Mulling. The proportions place it among ordinary, working Gospel books rather than among the deluxe altar books.

The text is laid out in a single column, normally of twenty-six lines to the page, ruled for a modest and functional presentation. The script is Insular, and here a point of some delicacy arises which will be addressed more fully below: the catalogues variously describe the hand as “Irish minuscule” and as “Irish majuscule,” and the most recent detailed textual study characterises it specifically as insular minuscule. This apparent inconsistency in the literature is not merely clerical — Houghton’s own catalogue entry for the manuscript in the standard Oxford guide describes the script as “Irish majuscule” while his fuller 2019 study calls it “insular minuscule” — and it reflects both the genuinely mixed grade of the writing and the differing conventions of the describers. The writing employs the characteristic device of diminuendo — the gradual diminution in the size of letters following an enlarged initial, so that the eye is led from a large opening letter down through progressively smaller forms into the body of the text. This is a hallmark of the Insular scribal tradition, found already in the earliest witnesses such as the Cathach attributed to the circle of Columba, and its presence in the Garland of Howth aligns the book firmly with that tradition.

The scribal execution is, by common assessment, workmanlike rather than distinguished. The manuscript has repeatedly been described as the work of several scribes, none of them of the first rank; the most recent study concludes that at least four copyists were involved. The text is divided into sense units by enlarged letters, punctuation is used only minimally, and words are separated by blank space, with prepositions often attached to the following word and with occasional unusual word-division across line ends. There is extensive abbreviation beyond the standard nomina sacra: the Insular symbols for autem, enim, and eius are used, together with a symbol for est, and two different abbreviations for quae. Corrections are few, and — a detail of some interest for the book’s later reception — there are no marginalia and no marginal apparatus such as the Eusebian canon references that structure many contemporary Gospel books.

The orthography, examined closely in the most recent textual study, presents a coherent if idiosyncratic system of Insular spellings — the interchange of long i and e, simplification of double consonants, erroneous duplications (nissi for nisi, occulos for oculos), interchange of o and u, and various treatments of the diphthong ae. Some features appear to be peculiar to this manuscript, such as the frequent writing of the third-person plural perfect ending -erunt as -iarunt. Alongside this systematic orthography there are frequent outright scribal errors: omissions through homoeoteleuton, dittography, and mistaken expansions of abbreviations. The cumulative impression is of scribes working from an Insular exemplar (a conclusion supported by tell-tale misreadings that only make sense if the source was in Insular script) but working without the highest care, so that readings unique to this book must be weighed critically before being accepted as genuine ancient variants rather than copying slips.

The State of Preservation and Conservation History

The manuscript survives only as a fragment of its original self. Its 86 leaves preserve portions of all four Gospels but with substantial lacunae, and the losses are unevenly distributed. The gaps span, among others, Matthew 27:59–28:20, Mark 5:31–6:13, and, in Luke, both 1:1–13 and 23:15–24:53; Mark is otherwise comparatively well preserved, while John is the most severely damaged of all, with only John 5:12 to 10:3 — roughly one-fifth of the Gospel — remaining extant. A number of leaves have suffered tears, holes, loss of margins, and discoloration. A telling codicological observation is that the back of the leaf on which the text of Matthew breaks off mid-verse is blank, which raises the possibility that the conclusion of that Gospel was never in fact copied — a hint of an incomplete or interrupted production.

For most of its modern scholarly life the manuscript was accessible only with difficulty and in a physically unsatisfactory condition. This changed decisively with the Early Irish Manuscripts Project undertaken at Trinity College Dublin — the Bank of America Merrill Lynch–funded Gospel Books conservation campaign, directed jointly by the Library’s Keeper of Preservation and Conservation, Susie Bioletti, and by the head of the College’s Department of History of Art and Architecture, Rachel Moss. Between roughly 2014 and 2017 four of Trinity’s early Gospel books — the Codex Usserianus Primus, the Garland of Howth, the Book of Dimma, and the Book of Mulling — were comprehensively conserved, technically examined, digitised, and studied art-historically. The four were publicly displayed together and made available online in high-resolution images for the first time; the digital Garland of Howth was launched in 2016. The conservation work included non-destructive scientific analysis of the pigments, and the digitisation had a direct textual benefit: portions that had been illegible to the manuscript’s early-twentieth-century editor could subsequently be read more clearly, enabling a fresh transcription.

The pigment analysis deserves specific notice because it is among the most concrete recent additions to knowledge of the book. Using X-ray fluorescence (XRF) supported by Raman spectroscopy, the conservation team examined the illuminated pages directly. On folio 22r, the opening page of Mark, the white pigment yielded a very strong calcium peak, and Raman analysis confirmed it as gypsum (calcium sulphate) — identified by a characteristic peak at 1005 cm⁻¹ — rather than chalk or lead white, the same white extensively used in the Book of Kells, particularly for hands and faces and mixed with indigo. This finding aligns with the broader Insular palette established by the same conservation tradition — the micro-Raman study of the Book of Kells identified blue as indigo, red-orange as red lead, yellow as orpiment, green as a vergaut mixture, black as carbon and iron-gall ink, and white as gypsum. Popular accounts describing the surviving illumination as “highly colourful,” with orange, yellow, white, and blue, are thus broadly corroborated by the scientific analysis, though the precise folio-by-folio distribution of pigments has not been exhaustively published.

The Decoration and Its Place in Insular Art

The Garland of Howth is, by the standards of the great Insular Gospel books, sparingly decorated, and its ornament survives only in part. Of an original programme that would presumably have furnished a decorated opening for each of the four Gospels, only two illuminated pages remain: folio 1r, the opening of Matthew, and folio 22r, the opening of Mark. The corresponding decorated openings for Luke and John are lost. This partial survival is itself an important qualification on any art-historical judgement: we are assessing a fraction of the original decorative scheme.

The Chi‑Rho Page of the Garland of Howth

The opening leaf of the Garland of Howth presents one of the most unusual solutions to the decorated Gospel incipit in the Insular repertoire, and one of the most difficult to read in its present condition. Unlike the great display pages of Kells or Lindisfarne, where ornament occupies a field cleared for it, here the decorative armature has been laid over an already written page: the lines of Insular script remain visible beneath and between the painted forms, so that letters and ornament share the same surface and compete for it. Whether this reflects a deliberate design or a decorative campaign superimposed on a text block already ruled and written is not established, and the abraded state of the leaf makes the question difficult to resolve from the image alone.

The composition is organised around a cross-shaped framework that divides the lower two-thirds of the page into rectangular compartments. Along the top runs a broad horizontal band of interlace terminating in zoomorphic heads with elongated jaws and spiralled snouts, from which ribbon-bodies descend to form the left border; a second, denser knot of interlace occupies the upper left corner, its crossing strands describing a saltire that is generally understood as the chi of the Chi-Rho monogram opening Matthew 1:18,

Christi autem generatio.

Down the right-hand edge, a vertical column of panelled display capitals — set in white and pale ochre against red-brown grounds, their forms rectilinear and articulated with lozenges and triangular voids — carries the continuation of the incipit. The reading of individual letters is uncertain at this resolution, but the arrangement is the familiar Insular device of running display script vertically down the margin when the ornament has consumed the horizontal space.

Within the cross-framed field, four figures are disposed in the quadrants. In the upper right compartment a winged figure inclines forward, its wing rendered as a lattice of small squared cells and its feathers as banded strips; a hand is extended toward the centre. The upper left compartment holds a corresponding form, largely reduced to an arched silhouette filled with vertical striations, which may once have been a second winged figure. Below, two seated frontal figures dominate. The one at lower left has patterned hair or a headdress worked in tight scrolls, a body clothed in obliquely striped drapery, and holds a long diagonal object; the one at lower right, with a smooth rounded head, grasps across its chest an object rendered with a hatched, blade-like shaft. The remaining ground is filled with further interlace, spirals, and beast heads, especially at the lower edge, where a serpentine body coils back on itself beneath the frame.

The identification of these figures remains genuinely open. There is broad agreement that the seated figure at lower left is likely to be Matthew; the figure holding the elongated object has resisted secure identification, and mid-twentieth-century specialists of Irish art frankly described the page as disconcerting. The most attractive proposal — that the four represent Abraham and David below with Isaiah and an angel above, thus supplying pictorially the genealogy of Christ that the ornamented page interrupts — should be presented as a hypothesis rather than as an established reading.

The surviving palette is consistent with what the non-destructive analysis carried out during the Trinity conservation campaign found elsewhere in the book: an opaque white identified as gypsum, a warm red-brown consistent with red lead, and traces of a darker blue-green in the compartment grounds and in the wing lattice. Much of what now reads as olive or grey is the product of abrasion and darkening rather than of original intention, and the page must have been considerably brighter and more legible in its contrasts.

The physical condition is, finally, part of what the image records. The leaf has lost its outer margins and is now inlaid within a paper or tissue support, visible as a lighter border at the gutter and along the lower edge. The vellum is heavily discoloured, browned at the fore-edge and stained across the lower left. Small holes interrupt the painted surface, and a long tear running diagonally through the lower right quadrant has been closed with coarse thread in a line of crossed stitches that remains conspicuous. Text from adjacent leaves shows through in places, adding to the visual congestion.

For an object that survived not as a working book but as a venerated relic — handled, carried, sworn upon — this cumulative damage is itself historical evidence, and it explains why the page was so long neglected by art historians before high-resolution imaging made its structure legible again.

The Initium Page of the Garland of Howth

The second of the two surviving decorated openings in the Garland of Howth introduces the Gospel of Mark, and it is by some margin the more legible and the more architecturally coherent of the pair. Where fol. 1r overwhelms its page with figures and interlace superimposed on the written text, fol. 22r deploys a monumental initial as the organising structure of the whole leaf, and the result is a design closer to the mainstream Insular solution for a Gospel incipit.

The composition is dominated by two immense vertical shafts rising through nearly the full height of the page, each drawn as a broad band with a pale core — the whitened inner strip is still clearly visible — flanked by darker outlines. Between them, at mid-height, runs a bold diagonal stroke. Taken together, these elements constitute the ligatured IN of Initium euangelii Iesu Christi, the opening words of Mark: the first shaft supplies the I, the diagonal and second shaft the N. This treatment of a Gospel opening as a single vast letter-frame, whose shafts function simultaneously as script and as architecture, is characteristic of Insular display pages, and the Garland here shows its makers working confidently within that idiom.

The upper terminals of both shafts expand into massive knots of ribbon interlace, symmetrically disposed and set against grounds of red-brown, with the strands drawn in a light ochre outlined in dark brown. Between and around these knots the ornament turns zoomorphic: at the head of the right shaft a beast with a long, open jaw and a prominent spiralled eye bites downward, its body dissolving into further ribbonwork, while at the upper right corner of the frame a second, smaller creature with a hooked beak turns back on itself. A comparable beast-head, with an elongated snout and lentoid eye, projects into the space between the shafts at roughly the level of the diagonal. At the foot of the page the ornament resolves into two large circular bosses of tightly wound spiral and interlace, the left one nearly free-standing, the right one caught in a broad band of knotwork that closes the design along the lower edge; from this band a further elongated beast, its jaw open and its neck extended, emerges towards the centre.

Set within the field enclosed by the letter, in a shallow arched niche formed by the interlace itself, is a small frontal figure. The head is drawn with disproportionately large, staring eyes, a long nose, and a small mouth, and is surrounded by an arched nimbus or head-covering; the shoulders and arms are indicated schematically, and the hands are brought together at the chest to hold a rectangular object, almost certainly a book or codex. Below the figure the drapery gives way to interlace terminals and a triangular panel of stepped ornament that reads as a stylised throne or lectern base. The identity of the figure is not securely established. A frontal, haloed, book-bearing figure at the head of Mark’s Gospel would most naturally be read either as the evangelist himself or as Christ; the schematic drawing offers no attribute that decides the question, and no lion symbol is discernible. Prudence suggests describing it as a frontal book-bearing figure of uncertain identity rather than asserting an evangelist portrait.

Down the right-hand side of the page runs a vertical register of panelled display letters, arranged in stacked compartments against red-brown and pale grounds, carrying the continuation of the incipit. Their forms are rectilinear, blocky, and heavily abraded; individual readings cannot be recovered with confidence from the image alone, though the disposition is the familiar Insular practice of running the remainder of the opening words vertically once the initial has consumed the page. The outer margin is closed by a narrow band filled with a dense, finely drawn fret or interlace, and a small square panel of lattice ornament terminates it at the lower right.

The palette is of particular documentary interest, since it is this folio that was subjected to non-destructive analysis during the Trinity College conservation campaign. X-ray fluorescence indicated a strong calcium signal in the white, and Raman spectroscopy identified it as gypsum — calcium sulphate — rather than chalk or lead white, the same white used extensively in the Book of Kells. The white survives most visibly in the cores of the two shafts and in the panels of the right-hand register. Alongside it the page preserves a warm red-brown, consistent with red lead, in the grounds behind the interlace, a muted blue-grey in several compartments, and a yellow-ochre for the ribbon strands themselves. Much of the present olive tonality is the product of abrasion, darkening, and the migration of the parchment’s own colour rather than of original design.

Later hands have left their traces at the head and foot of the leaf. At the top, above the ornament, a post-medieval annotator has written Cap. I: together with a further, partly illegible inscription; at the bottom appear the modern folio number 22, a separate figure 43, and an additional pencil or ink notation. These interventions belong to the manuscript’s early modern and modern custodial history and are a reminder that the book was catalogued, foliated, and repeatedly renumbered long after it had ceased to function as a working Gospel book.

The physical condition, finally, is markedly better than that of fol. 1r but still compromised. The leaf has been inlaid, and the join is visible as a lighter strip along the gutter and the lower edge; the outer margins are irregular and torn, with losses along the bottom left; the parchment is browned and stained, particularly towards the fore-edge; and faint show-through of text from the verso and from adjacent leaves is discernible within the shafts. On the facing leaf, at the left of the image, a hole in the parchment is plainly visible. That so much of the ornamental structure remains legible despite these losses is what has allowed the page to serve as the principal evidence for the artistic ambitions of a book long dismissed as merely workmanlike.

The art-historical significance of the manuscript lies precisely in its ordinariness and its idiosyncrasy. It was, until the recent conservation project, “almost unknown to art historians,” and one of the project’s stated aims was to reassess it from an art-historical point of view for the first time. What the reassessment suggests — on the securely sourced reading of the specialist chapter that resulted from it — is that the scribes and artists of the Garland of Howth drew on a considerable array of exemplars, of varied type and varied date. In other words, the book is a compilation of models: its makers had access to, and combined, sources of different periods and characters. This is a valuable insight into the working conditions of an Insular scriptorium beyond the first rank, and it repositions the Garland of Howth not as a failed attempt at Kells-like grandeur but as evidence for the wide circulation of scripts, materials, and techniques within the Insular world.

The Old Latin Text and Its Textual Affiliations

If the decoration of the Garland of Howth is modest, its text is of the first importance, and it is as a textual witness that the manuscript has earned its place in the scholarly literature. The book contains the four Gospels in the standard Vulgate order (Matthew, Mark, Luke, John), but its text is emphatically not a pure Vulgate. It is a “mixed” text, and the nature of the mixture is precisely what makes it valuable.

The foundational observation, established by Abbott in his 1884 collation and refined by subsequent scholars, is that the manuscript exhibits what textual critics call “block mixture”: different parts of the book descend from different kinds of exemplar. The text of Matthew is substantially Old Latin — that is, it preserves readings from the pre-Vulgate Latin translation of the Gospels that circulated before Jerome’s late-fourth-century revision — whereas the texts of Mark, Luke, and John are close to the Vulgate. Early study located the Old Latin element in Matthew, and a refinement soon established that the Old Latin section of Matthew does not begin at the start of the Gospel but only partway through, from about chapter 16 onward, with the earlier chapters closer to the Vulgate. This phenomenon of block mixture — the joining within a single manuscript of textually distinct sections copied from distinct sources — is itself an important datum for understanding how Gospel texts were transmitted and assembled in early medieval Ireland.

The significance of the Matthew text can be quantified, thanks to the computer collations of the late twentieth century and their re-analysis in the most recent study. Measured against the editorial text of the modern critical Stuttgart Vulgate, the Garland of Howth’s overall agreement across all its collated passages is 78.9%, which — in Houghton’s words — places it among the twenty witnesses which differ most from this standard form.

But the figures for individual passages of Matthew are more dramatic still. Across the collated test passages the agreement ranges from 85.8% in a passage of Luke down to just 56.5% in a test passage drawn from Matthew 26 and 27 — where, in the words of the most recent study, the manuscript exhibits the greatest difference from the Vulgate of all Latin gospel manuscripts copied in the first millennium.

This is a striking statistic, and it must be handled with the care the recent analysis itself urges: the “competition” in that passage is reduced by the fact that two of the oldest and most divergent Old Latin witnesses do not survive for it. Even so, the divergence exceeds that of celebrated Old Latin codices, and it establishes the Garland of Howth as a genuinely important, if long neglected, Old Latin witness in Matthew.

The manuscript’s closest textual relatives, on the evidence of the collations, are also instructive. In the Old Latin portion of Matthew its readings align most closely with the Codex Usserianus Primus (VL 14) and with the Codex Claromontanus (VL 12), two manuscripts often characterised as sharing a “Gallo-Irish” text whose ultimate origins may lie in fourth-century Italy. The overlap with these two witnesses is marked: a substantial number of readings are shared with one or both of them and with no other surviving Old Latin manuscript.

There are also cases in which a reading of the Garland of Howth is shared, among all surviving first-millennium Latin manuscripts, only with the Codex Usserianus Primus, underlining the special relationship between the two Usserian codices at the level of text as well as of shelving. The manuscript also preserves Old Latin interpolations known from other witnesses — for example at several points in Matthew — and even a long interpolation at the end of Matthew 27:49, drawn from John’s account of the piercing of Christ’s side, which, though unparalleled in the directly transmitted Old Latin tradition, appears in Greek majuscules and thus testifies to the antiquity of the reading.

It should be emphasised, to correct another loose statement sometimes met with, that the Garland of Howth is not uniformly an “Old Latin manuscript.” Its Old Latin character is concentrated in the latter half of Matthew. In Mark and much of Luke it is a mixed text with a higher proportion of Vulgate readings, and in John it is essentially Vulgate — so much so that the editors of the modern critical Vetus Latina edition of John assigned it to a group whose affiliation is with the Vulgate (with a few non-standard readings shared with other Irish witnesses), while the editors of the Vetus Latina edition of Mark decided to exclude it altogether on the grounds that its text there is Vulgate. The phrase “Vetus Latina” as applied to this book therefore requires qualification: it is a legitimate and valuable Old Latin witness, but principally for Matthew, and its overall complexion is that of a mixed Irish Gospel text in which Old Latin survivals persist — as they characteristically did in Ireland — later than in most of Latin Christendom.

The History of Textual Scholarship on the Manuscript

The modern study of the manuscript’s text has a well-defined lineage that is worth setting out, both for its intrinsic interest and because it illustrates how a witness can be at once repeatedly edited and persistently neglected. The first systematic engagement was the collation published by T. K. Abbott in 1884 as an appendix to his edition of the Codex Usserianus Primus, a volume whose full Latin title (Evangeliorum versio antehieronymiana ex codice Usseriano … adiecta collatione codicis Usseriani alterius) announces its programme of presenting the pre-Hieronymian (pre-Jerome) Gospel version of the Usserian codex with a collation of the “other” Usserian codex appended, using the Codex Amiatinus as the base text and marking agreements with the Codex Usserianus Primus and the Book of Kells. It was Abbott, in effect, who introduced the manuscript into the mainstream of biblical scholarship and whose collation was used to cite it in the Oxford Vulgate.

A few years later, in 1897, a further contribution refined Abbott’s conclusions by observing the precise point at which the Old Latin section of Matthew begins, and published a full transcription of the Old Latin portions as an appendix to a study of the Book of Mulling — situating the Garland of Howth within the broader phenomenon of block mixture in Insular Gospel books.

The most substantial edition came in 1919, when H. C. Hoskier published his edition of the surviving portions of the manuscript, The Text of Codex Usserianus 2. R2. (“Garland of Howth”). Hoskier’s work is what has been described as a “textual facsimile”: the transcription matches the layout of the manuscript and reproduces, through special characters, the appearance of the Insular script, the abbreviation symbols, and the forms of particular letters.

It is typographically impressive and remains genuinely useful for reading damaged passages, but the wholesale preservation of the abbreviations makes it a difficult instrument for the non-specialist. Hoskier was scathing about his predecessor’s accuracy, characterising Abbott’s collation as “deplorable” and cataloguing what he claimed were a thousand errors of omission and commission and a failure to report over two thousand varieties of spelling.

His own transcription is generally very accurate, though it too has been shown to be improvable in places — famously, he missed the opening words of Matthew 1:18 on the first surviving page, an oversight later noticed. Hoskier’s introduction, however, confined itself to listing a dozen or so readings of interest per Gospel, several of them marshalled in support of his own theory that the Latin version originated in a bilingual manuscript; he offered no analysis of the book’s position within the Latin textual tradition or of its block mixture.

After Hoskier the manuscript entered a long period of relative neglect. It was included in the inaugural register of Old Latin manuscripts issued by the Vetus Latina Institute in 1949 and assigned the number 28, but — unlike the manuscripts on either side of it in the numbering — it was not excerpted onto the index cards that formed the basis of the Vetus Latina Database, and it was not included in the volumes of the Itala edition of the Old Latin Gospels. The result was that a genuinely important witness fell through the cracks of the standard reference tools. Its text was not seriously re-examined until the computer collations of substantial test passages across more than 450 Latin Gospel manuscripts undertaken in the 1980s, whose statistical tables — published only posthumously — first revealed just how divergent the Garland of Howth’s Matthew text is.

The most recent and most thorough analysis, published in 2019 by H. A. G. Houghton, brought the manuscript into the digital age. Drawing on the unpublished data from the earlier computer collations together with a fresh full-text electronic transcription of Matthew — made from the new high-resolution images produced by the Trinity conservation project — Houghton’s study offered a “born-digital” reassessment that confirmed and quantified the manuscript’s Old Latin affiliation in the latter half of Matthew, its close relationship to the Codex Usserianus Primus and the Codex Claromontanus, and the presence within it of readings of real text-critical value. This study is the essential modern point of departure for anyone approaching the manuscript’s text, and it explicitly frames the Garland of Howth as a “neglected Old Latin witness” deserving of the attention it had for so long been denied. A complementary fresh transcription of Matthew, produced by a related Irish Gospel books project, is also available.

Palaeography and the Dating Debate

The dating of the Garland of Howth is a matter on which precision is impossible and on which the sources display a genuine, and instructive, spread. The range of dates encountered in the literature runs from the eighth century to the tenth. Several strands must be distinguished.

The palaeographical placement rests above all on Lowe’s “Codices Latini Antiquiores”, where the manuscript is catalogued as CLA II.272 and dated to the eighth or ninth century (saec. viii/ix). This is the anchor point for scholarly dating. Building on this, the most recent textual study describes the book as copied in Ireland “around the year 800” and as a “ninth-century” Gospel book, and this — the late eighth to early ninth century, centring on circa 800 — represents the current scholarly consensus and is the dating that should be preferred. The Trinity conservation and art-history team likewise consistently described the manuscript as of the ninth century (or, more cautiously, the eighth-to-ninth century).

A wider range is met with in older and in popular literature. The encyclopaedic tradition frequently gives “8th to 10th century” as a bracket, and a number of accessible sources state that the book is “believed to date from the late ninth to early tenth centuries.” These later datings are not baseless — they reflect the workmanlike and in some respects retardataire character of the script and ornament — but they are looser than the palaeographical evidence strictly warrants, and the “around 800” / ninth-century placement is better supported.

One claim in particular must be addressed and firmly qualified, because it has begun to circulate. Some tertiary accounts assert that the art-historical reassessment by Rachel Moss “refines the dating to the eleventh century” and describes the manuscript as “polytemporal.” This assertion should be treated as unreliable as stated. It appears in an AI-generated compendium without a controllable citation to Moss’s actual chapter, and it is internally contradicted even within that same source, which elsewhere dates the manuscript to the late eighth or early ninth century.

What the securely documented version of Moss’s argument actually maintains — on the evidence of a peer-reviewed review of the volume in which her chapter appeared — is quite different and more subtle: that the scribes and artists of the Garland of Howth drew on exemplars of varied type and varied date, so that the book must be understood as combining models of different periods. This is a statement about the mixed chronological character of the sources the makers used, not a claim that the manuscript itself was produced in the eleventh century.

The distinction matters, and conflating the two produces exactly the kind of misattributed date this dossier is meant to correct. Until Moss’s chapter can be consulted directly, the responsible position is that the manuscript was made around 800, in the ninth century, by scribes drawing on a chronologically heterogeneous body of exemplars.

A further, subtler point of palaeographical description also deserves airing. The catalogues disagree, at least verbally, over whether the script is “majuscule” or “minuscule”: some descriptions (including the entry in the standard Oxford catalogue of Latin New Testament manuscripts, which calls it “Irish majuscule,” and various database records) use majuscule, while the most recent close study calls it insular minuscule.

This is not merely a clerical discrepancy; it reflects the mixed grade of the writing and the difficulty of forcing an Insular hand of the second rank into the tidy categories devised for the calligraphic masterpieces. The honest summary is that the book is written in an Insular hand of mixed grade, employing diminuendo and the full apparatus of Insular abbreviation, executed by several scribes of moderate competence.

Association with St Nessan, Ireland’s Eye, and the Question of Origin

The traditional account of the manuscript’s origin ties it to the small monastic settlement on Ireland’s Eye (Inis Mac Neasáin), the rocky islet immediately north of Howth Harbour. According to this tradition, a monastery was founded there — the standard account attributes it to the sons of a certain Nessan around the early eighth century, or, in some tellings, places St Nessan’s foundation in the later sixth century — and it was in the scriptorium of this island monastery that the Garland of Howth was written. The book was subsequently kept in the parish church on the mainland at Howth. The island’s Irish name, Inis Mac Neasáin (“island of the sons of Nessan”), and the dedication of its ruined church to the “Three Sons of Nessan” preserve the association, which also appears in a Latin form (Insula Filiorum Nessani) in a papal document of the twelfth century.

This origin story is traditional and plausible but not strictly proven, and several layers of caution are required. First, and most importantly, the identity of the “St Nessan” involved is itself contested. Later scholarship showed that the attribution of the foundation to a “St Nessan” rests in part on misinterpretation and on the confusion of two men bearing that name — one of them a saint from another part of Ireland with no genuine connection to Ireland’s Eye. The nineteenth- and early-twentieth-century antiquarians who worked on the site disagreed about the genealogy and floruit of the founder or founders, some placing the “sons of Nessan” in the late seventh century, others reasoning from supposed associations with known saints. The upshot is that the hagiographical framework of the origin tradition is insecure.

Second, the specific claim that the manuscript was physically written on Ireland’s Eye is a reasonable inference from its later custody there, but it is not demonstrable from the book itself, which contains no colophon or scribal note naming a place of origin. The most that the physical evidence supports is that the book was made in Ireland — the Insular script, the Insular exemplar behind it, and the incorporation of an Irish-language gloss all confirm an Irish milieu — and the most cautious recent framing describes it simply as copied in Ireland around 800, associated with (rather than certainly produced at) the monastery on Ireland’s Eye. The local tradition remains the best available account of where the book was made and kept, but it should be presented as tradition supported by circumstantial evidence, not as an established fact.

The Irish-language gloss just mentioned is itself a small but telling piece of evidence for the Irish origin. At one point in Matthew the copyist, evidently misunderstanding his exemplar, wrote a garbled form of an Irish translation of the phrase “the money that was taken” (airgid gabtha) in place of the Latin — an incorporated gloss that betrays both the Irish linguistic environment of the scriptorium and the fact that the exemplar was itself an Insular book.

The Later History: Relic, Legend, and Swearing-Book

The post-medieval history of the Garland of Howth belongs as much to the study of popular devotion and local custom as to codicology, and it exemplifies a phenomenon well established for early Irish Gospel books: their survival not primarily as texts to be read but as sacred objects, relics venerated for their association with a founding saint. It has been observed of the great Irish Gospel books generally — Durrow and Kells among them — that they owe their survival to precisely this special status as relics, often enshrined in decorated metal book-shrines (cumdachs). The Garland of Howth fits this pattern: kept in the parish church at Howth, it functioned less as a working Gospel book than as a venerated heirloom of the island monastery’s sanctity.

Around it accreted legend. The most colourful is the story, reported in nineteenth-century antiquarian collections of local lore, that one of the sons of Nessan (or St Nessan himself) hurled the holy book at the devil during a temptation on Ireland’s Eye, and that the force of the throw flung the devil against the mainland cliff, splitting the rock and creating the feature known as Puck’s Rock. Such stories, in which a saint’s book becomes a weapon against demonic assault, are a recognisable type in Irish hagiographical folklore, and they attest to the manuscript’s standing as a numinous object charged with protective power.

Connected to this sacral status is the tradition that the book served as a “swearing-book” — an object upon which oaths were sworn, its sanctity guaranteeing the truth of the swearer. Oath-taking on venerated Gospel books and other relics is amply documented in medieval and early modern Ireland, and the reputation of the Garland of Howth as such an object is part of the local memory attached to it. The reader should note, however, that the detailed testimony for this use comes largely from antiquarian report rather than from contemporary medieval record, and the specifics should be treated as tradition. What is secure is the general point: the manuscript was preserved and prized for centuries as a holy relic of the Howth locality, and it is to that veneration, not to any scholarly interest, that we owe its survival at all.

Provenance and Arrival at Trinity College Dublin

The manuscript’s journey from the church at Howth into the Library of Trinity College Dublin is less well documented than one would wish, and the gaps should be acknowledged rather than papered over. The broad outline is clear: having been kept at Howth through the later Middle Ages and into the early modern period — where, as noted, it was recorded under a form of its Irish name in the early sixteenth century — it eventually entered the Trinity collections, where it received its “Usserianus Secundus” name in honour of Archbishop Ussher and was shelved alongside the Codex Usserianus Primus.

The Ussher connection provides the most likely mechanism. Ussher was the greatest Irish book collector of his age, dispatched in his youth to purchase books for the newly founded university and thereafter assembling one of the finest libraries in the British Isles. After his death in 1656 his library had a complicated fate: Cromwell prevented its export, and it was eventually acquired — through a sum subscribed by the army in Ireland — as a donation intended for Trinity College.

The books were for a time held at Dublin Castle, exposed to loss, and only after the Restoration was the surviving remnant transferred to the College Library. It is within this general movement of Ussher’s collection into Trinity that the two Usserian codices are understood to have arrived, though — as with the Primus — one should be wary of assuming that the “Ussher” name proves personal ownership in a strict legal sense.

The precise details of when and how the Garland of Howth passed from Howth to Ussher (or to Trinity), and the identity of any intermediate custodians in the seventeenth and eighteenth centuries, are not documented in the surviving records, and honest accounts admit as much. What can be said securely is that the manuscript has been a Trinity manuscript since the seventeenth century, catalogued as MS 56 with the older shelfmark A.4.6, and that it takes its place there within the pre-eminent collection of early Irish Gospel books that includes Kells, Durrow, Armagh, Dimma, Mulling, and the Codex Usserianus Primus.

Comparison with the Codex Usserianus Primus

Because this essay is conceived as a companion to a dossier on the Codex Usserianus Primus, an explicit comparison of the two is in order, and it is illuminating precisely because the two books are so often confused and yet are so different.

The Codex Usserianus Primus (MS 55; r1 / VL 14 / CLA II.271) is the older and, textually, the purer of the two. It is generally dated to around the early seventh century — some scholars would place it as early as the late sixth century, and its origin has been debated between Ireland and the Columban foundation of Bobbio in northern Italy, with the current view favouring Irish manufacture and holding it possibly the earliest surviving Irish codex.

It is a thoroughgoing Old Latin Gospel book of the European type, with the Western order of Gospels (Matthew, John, Luke, Mark), textually notable for such features as its naming of one of the thieves crucified with Christ. Its decoration consists of a single surviving ornamented leaf: a cross outlined in dots, flanked by the Greek letters alpha and omega, standing between the explicit of Luke and the incipit of Mark within a framed field — the earliest surviving instance of an Insular artist copying a Mediterranean decorative form, and thus a document of the very beginning of the Insular illuminated tradition. Physically it survives as fragments of roughly 180 leaves, discoloured and remounted on paper.

The Garland of Howth (MS 56; r2 / VL 28 / CLA II.272) is, by contrast, some two centuries later — around 800 — and its text is mixed rather than purely Old Latin, with its Old Latin character concentrated in the latter half of Matthew and its remaining Gospels close to the Vulgate. Its order of Gospels is the standard Vulgate order. Its decoration, though also fragmentary, is more ambitious in conception than the Primus’s single dotted cross: it retains two figured and ornamented Gospel openings with display lettering, interlace, and enigmatic figural imagery. Where the Primus stands at the threshold of the Insular tradition, copying an eastern cross with dotted outlines, the Garland stands within a mature Insular idiom, deploying diminuendo, interlace initials, and a coloured palette.

Yet the two are genuinely linked, and not only by their shared shelf and their common Ussher name. At the level of text they are relatives: in the Old Latin portion of Matthew the Garland of Howth agrees more closely with the Codex Usserianus Primus than with any other surviving witness, and certain readings are shared, among all first-millennium Latin manuscripts, by these two alone. Both belong, in that portion, to the “Gallo-Irish” strand of the Old Latin tradition. The relationship is thus real but partial: the two Usserian codices are cousins in the Old Latin Matthew, but they diverge sharply in date, in the purity and extent of their Old Latin text, in their order of Gospels, and in their decoration. To treat them as interchangeable — as careless accounts sometimes do — is to lose exactly the information that makes each valuable.

Significance for Insular Book Production and the Transmission of the Old Latin Gospels in Ireland

The broader importance of the Garland of Howth can be summarised under two heads, corresponding to the two disciplines that have claimed it.

For the textual history of the Latin Bible, the manuscript is a witness of the first importance to the survival of Old Latin readings in Ireland long after the Vulgate had become standard elsewhere. Its block mixture — Old Latin Matthew grafted onto largely Vulgate Mark, Luke, and John — is a concrete illustration of how Insular scriptoria assembled their Gospel texts from heterogeneous exemplars, and its extraordinarily divergent Matthew text, the most non-Vulgate of any first-millennium Latin Gospel manuscript in certain passages, preserves ancient readings that would otherwise be lost or attested only in Greek.

Its close kinship with the Codex Usserianus Primus and the Codex Claromontanus helps to define the “Gallo-Irish” text-type and to trace a line of transmission running, in the view of the specialists, from fourth-century Italy through Gaul into Ireland. That such a witness could remain effectively unstudied between 1919 and the twenty-first century — omitted from the Vetus Latina Database and the Itala, excluded from the critical edition of Mark, and only belatedly reincorporated through digital methods — is itself a cautionary tale about the fortunes of manuscripts, and its recent recovery is a model of how conservation, digitisation, and textual scholarship can combine to rehabilitate a neglected source.

For the history of Insular art and book production, the manuscript is valuable precisely because it is not a masterpiece. The overwhelming attention lavished on Kells, Durrow, and Lindisfarne can distort our picture of Insular book culture into a gallery of supreme achievements. The Garland of Howth restores balance: it is the kind of workmanlike, second-rank Gospel book that must have been far more numerous than the survivors suggest, produced by several moderately skilled scribes drawing on a miscellany of exemplars of varied date and type.

Its very idiosyncrasies — the puzzling figural page, the mixed grade of script, the incomplete production hinted at by the blank leaf after Matthew, the retardataire clinging to Old Latin readings — make it a rich document of ordinary Insular practice. The recent art-historical reassessment, which rescued it from being “almost unknown to art historians,” suggests that its makers had access to a considerable range of models, and thereby uses this modest book as evidence for the wide circulation of scripts, materials, and techniques across the Insular world. In this sense the Garland of Howth is significant not despite its obscurity but because of it: it lets us glimpse the broad base of the pyramid whose apex is the Book of Kells.

Conclusion: What Is Known, Debated, and Unknown

The Garland of Howth can be characterised, in closing, according to the epistemic layering that has structured this essay. Securely established are the physical facts: a parchment Gospel book of 86 surviving leaves, about 241 × 192 mm, one column of some twenty-six lines, written in an Insular hand of mixed grade by at least four scribes, with two surviving decorated Gospel openings, catalogued at Trinity College Dublin as MS 56 (A.4.6), designated r2 / VL 28 / CLA II.272, conserved and digitised in the mid-2010s, and analysed pigmentwise (gypsum white on the Mark page, and a palette including indigo, red lead, and orpiment). Securely established, too, is the textual profile: a mixed text with a strongly Old Latin Matthew (from about chapter 16), close to the Codex Usserianus Primus and the Codex Claromontanus, and otherwise close to the Vulgate, edited by Abbott (1884) and Hoskier (1919) and re-analysed by Houghton (2019).

Scholarly consensus, slightly less secure but well founded, holds that the book was made in Ireland around 800 (eighth-to-ninth century), that its name “Garland” derives from the Irish Ceithre Leabhair (“four books”), and that it survived because of its status as a venerated relic. Genuinely debated, or at least imprecise, are the exact date within the eighth-to-tenth-century bracket, the correct verbal classification of its script (majuscule versus minuscule), and the identification of the figures on its Matthew page. Matters that remain unknown or merely traditional include the precise place of writing (Ireland’s Eye is tradition, not proof), the identity and historicity of the “St Nessan” of the foundation legend, the detailed custodial history between Howth and Trinity, and the specifics of its use as a swearing-book. Firmly to be rejected are several repeated errors: the conflation of the Garland of Howth with the Book of Howth; the treatment of the two Usserian codices as interchangeable; the misassignment of their sigla; and the confidently stated but unsupported claim that the manuscript has been re-dated to the eleventh century.

The Garland of Howth thus rewards exactly the kind of layered, source-critical attention that its long neglect denied it. It is a modest book with an immodest textual importance, a relic that outlived the community that made it, and a witness whose recovery in our own time — through conservation, digitisation, and patient collation — demonstrates that even the most overlooked survivors of the Insular world still have a great deal to teach.