Around AD 800, monks in Ireland or northern Britain created the Book of Kells, a Latin Gospel manuscript filled with painted initials, animals and intricate interlace. Its 340 surviving leaves contain the four accounts of Jesus’s life attributed to Matthew, Mark, Luke and John. It also includes introductions and tables that allow readers to compare passages across the Gospels. Today it is Trinity College Dublin’s manuscript MS 58, but its elaborate pages belong to the world of early medieval monasteries, where copying Scripture was a demanding religious and scholarly undertaking.

The Book of Kells is an illuminated manuscript, meaning that its handwritten text is accompanied by painted decoration. Its makers wrote on prepared calfskin, using carefully shaped letters alongside human figures, animals, interlaced patterns and geometrical designs. Making such a book required skilled scribes and artists, expensive materials and sustained work. The artists devoted unusual attention to the appearance of the Gospel text, especially at the openings of important passages. The book was probably produced for important church occasions, when a sacred manuscript could be displayed and venerated as well as read.

The makers left no inscription identifying themselves, and there is no contemporary record of the place where they worked. Many scholars favour Iona, the monastic island off Scotland’s western coast, although the community at Kells in Ireland preserved the manuscript in later centuries and supplied its modern name. Other locations have been proposed. The question remains open, and scientific study of the pages has also corrected long-repeated assumptions about how their colours were made.

Iona, Kells and the monks who could have made the book

Iona Abbey monastery buildings on the island of Iona
Iona Abbey on the island of Iona, traditional site of St Colum Cille’s monastery. Source: Wikimedia Commons

Colum Cille, known in Latin as Columba, founded the monastery on Iona in the sixth century. From this island in the Inner Hebrides, the community developed a network of associated monasteries across Ireland and northern Britain. Its monks studied and copied Christian books, maintained contact with other churches and taught the Latin literacy needed for such work. The Trinity College Dublin manuscript catalogue dates the Book of Kells to about AD 800 and reports that many scholars favour Iona as its place of manufacture. No surviving record identifies Iona as the place where the manuscript was made.

Scholars call the manuscript’s script and art Insular, a term for the distinctive traditions of book-making that developed in early medieval Ireland and Britain. The artists drew on the visual language of local metalwork and stone carving while working within Latin Christianity, which supplied the Gospel texts, the four evangelists’ customary symbols and models for arranging them on a page. The Lindisfarne Gospels, made in Northumbria about a century earlier, offer a particularly useful comparison. Their decorated initials, cross designs and portraits show how related artistic practices circulated well beyond any one monastery.

A monastery capable of making such a book needed far more than people who knew Latin. Calves had to be raised or skins acquired, hides prepared, inks and pigments supplied, and scribes and painters fed while they spent their days on work that produced neither crops nor ordinary goods for exchange. Its scale suggests that the community responsible could commit substantial labour and materials to a single Gospel book. It does not tell us whether every stage happened in one workshop or whether the skins and colours all came from the same place.

Raids on Iona around the beginning of the ninth century complicate the manuscript’s history. The island was attacked in 795 and 802, and the Annals of Ulster record the killing of 68 members of its community in 806. Soon afterwards, the Columban monks established an important refuge at Kells in County Meath. As Maldonado, Campbell, Clancy and Forsyth explain from annalistic and archaeological evidence, Iona continued to operate after the monks established a community in Ireland. The manuscript could have moved with the community, but no contemporary account says that it did. Nor can its unfinished decoration securely be explained as work interrupted by a particular raid.

A different proposed origin appeared in The Book of Kells: Unlocking the Enigma, published in 2025. Victoria Whitworth links the Book of Kells with Pictish eastern Scotland, and particularly with Portmahomack on the Tarbat peninsula. Excavations there found evidence for a monastic workshop in which animal hides were probably prepared for parchment. The Scottish Archaeological Research Framework describes the material traces, including equipment used for stretching skins and evidence of lime production. The Pictish workshop could have supported sophisticated manuscript production, though no archaeological evidence connects it specifically to MS 58. Whitworth’s proposed attribution remains disputed.

How the monks made the vellum pages

Madonna and Child folio painted on vellum
Madonna and Child folio painted on fine vellum. Source: Trinity College Dublin

The manuscript now contains 340 folios, or leaves, each with two usable sides. That gives 680 sides bearing writing or decoration. Trinity lists the leaves at approximately 330 by 255 millimetres, but they were originally larger. Their edges were cut back during nineteenth-century rebinding and gilded at the same time. Scholars think about 30 additional leaves have been lost. Some may have contained major illustrations, though no early complete list of the contents survives to show precisely which pages disappeared.

These leaves are vellum, a fine parchment made from animal skin rather than paper. Preparing a skin involved cleaning it, removing the hair, stretching it over a frame and scraping it repeatedly until it was thin and even enough for writing. As it dried under tension, its fibres became firm enough to accept ink on both surfaces. This was not the same process as tanning leather. A scrape made too deeply might cut through a skin, while an uneven area or natural flaw could obstruct a line of text. Marks and defects still visible in the Book of Kells make the preparatory labour tangible.

The scribes worked with gatherings of folded sheets rather than assembling a book from hundreds of unrelated loose pages. A skin folded to create two joined leaves formed a bifolium; several folded sheets could be nested to make a quire. These gatherings were later sewn together. The arrangement could affect the order of work, especially if a large illustration was prepared on a separate leaf while text was being copied elsewhere. It also left individual leaves vulnerable to detachment and loss during later rebinding.

How many calves supplied the book depends partly on whether one counts only the pages that survive or also those believed lost. In 2025 Trinity art historian Rachel Moss gave an estimate for the surviving pages at about 159 skins. In a separate study of the manuscript’s original extent, she cited Denis Casey’s estimate of at least 185 calves for the complete book. Neither is a figure preserved in a medieval account book. The number of usable leaves varied with the size of each hide and the way it was cut. The estimates certainly do not substantiate the old claim that one monastery maintained a herd of 1,200 cattle for the project. Skins might have arrived over many years from several estates.

Before writing began, the makers marked out the space available on each prepared sheet. Small punctures guided ruled lines for text and helped establish the boundaries of elaborate painted panels. The surviving marks show that the designs rested on measured preparation, even where the finished decoration appears almost impossibly free-flowing. We can inspect those traces, but no inventory of this particular workshop survives to identify every knife, frame or ruling instrument the monks employed.

Advertisements

The pigments: blue from plants, yellow from arsenic

For many years, descriptions of the Book of Kells claimed that the deep blue on its pages came from ultramarine, prepared from lapis lazuli brought from Central Asia. Such a finding would have implied access to an exceptionally costly pigment transported over great distances. Scientific examination has not supported the ultramarine identification. Susan Bioletti and colleagues used micro-Raman analysis at 681 sites across the manuscript’s four volumes in a 2009 study. They identified indigo as the blue colourant, consistent with dye from plants such as woad. Trinity consequently considers it unlikely that lapis lazuli was used in the manuscript.

The same investigation identified red-orange red lead, yellow orpiment, black pigments from carbon and iron-gall inks, white gypsum and purple associated with orcein derived from lichens. Brenda Doherty and colleagues carried out a second non-invasive investigation published in 2013 using additional spectroscopic techniques to study colourants and binding materials. They also identified anhydrite, a calcium sulphate mineral, and clarified the contribution of organic purple dyes. The studies tell us which materials survive on the pages. They cannot recover every medieval recipe or trace each ingredient to the place where it was obtained.

Orpiment, a mineral containing arsenic sulphide, gave the artists an intense yellow that could suggest the brilliance of gold. There is no evidence, however, that the original painted pages were decorated with applied gold leaf. Gold mentioned in the manuscript’s medieval history belonged to its valuable exterior covering; the gilding around today’s page edges is a nineteenth-century addition. Orpiment and red lead were hazardous substances, but identifying them does not tell us whether the painters were poisoned or what precautions they took.

According to Trinity, the copper-based green pigment reacted with moisture and damaged the calfskin in places, sometimes eating through the page. The artists also applied colours in layers, occasionally building up several applications to achieve stronger contrasts and richer surfaces. This careful use of paint explains much of the manuscript’s depth of colour without requiring the gold leaf familiar from many later illuminated books.

Even when a laboratory can identify a colourant, its route to the manuscript is harder to establish. Indigo could come from different plants, and orpiment alone cannot reveal which traders or monasteries supplied it to Iona or another workshop. Earlier descriptions named Mediterranean plants for purple, imported insects for some reds and egg white as the binding medium. The published studies do not securely identify those particular sources or materials for the Book of Kells. The combination of mineral pigments, organic dyes and layered paint required specialised preparation, though the routes by which individual materials reached the artists remain uncertain.

Writing the Gospels and planning the illustrations

Evangelist John with quill and ink at a writing desk
St John seated with quill, writing board, and inkpot, showing the tools of an insular scribe. Source: Trinity College Dublin

The main text is written in Insular majuscule, a rounded formal script whose clear letterforms could also accommodate decoration. Its Latin Gospels draw principally on the biblical text associated with Jerome, the late fourth-century scholar who translated and revised Scripture, but some readings come from older Latin versions. Prefaces and summaries precede the Gospels, together with canon tables. These tables set corresponding passages from the four evangelists alongside one another, allowing a reader to locate different accounts of the same episode.

Most of the writing was done in dark ink, with coloured letters appearing in selected places for emphasis. The portrait of St John shown above offers a medieval image of a scribe with writing equipment, but it represents a sacred author, not an eyewitness record of the workshop that produced the Book of Kells. We cannot use it to identify the bird that supplied a particular quill, establish that the artists used a cow-horn inkpot or reconstruct the precise form of their original binding.

Artists preparing the great painted initials needed to coordinate letters, panels, figures and interlaced lines before applying colour. A loop had to pass correctly over and under the next, and an animal’s body might also form part of a letter. Trinity’s Rachel Moss describes tiny grids of pinpricks beneath complex designs, including those beneath the famous Chi Rho page. The pinpricks record precise geometric planning. They provide no direct evidence for the frequently repeated suggestion that the painters used magnifying crystals or lenses.

Specialists have traditionally identified four main scribal hands and about three artists behind the manuscript’s most elaborate pages. A hand is the pattern of letter shapes, spacing and habits that permits scholars to distinguish one writer’s work from another. The Trinity catalogue adopts four scribes as a working assessment, while acknowledging that it is much harder to separate the work of copyists from that of painters. Whitworth’s 2025 study, among other recent scholarship, asks how securely such individual assignments can be made.

The same person might have written a passage and painted its surrounding letters, or several specialists might have returned to the leaf at different stages. Even seemingly unfinished work is open to interpretation. Benjamin Tilghman has suggested that the ornament surrounding Matthew’s genealogy on folios 29v to 31r was intentionally designed to look unfinished as a commentary on the biblical passage. Other scholars dispute that reading. Neither the production schedule nor the order of writing and decoration is known for every page, and physical evidence cannot supply the artists’ names, personalities or exact division of tasks.

The text itself contains mistakes. Trinity notes that a passage copied twice on folios 218v and 219r was corrected by placing small red crosses across the first version. Other sections omit letters or words, perhaps when a copyist looked ahead to the wrong place in an exemplar. Even carefully decorated passages could contain mistakes in the copied Latin. Their presence strengthens the view that the manuscript was intended chiefly for ceremonial use, rather than being the most convenient book for a monk who needed to consult an exact Gospel text every day.

Advertisements

The pages that made the Book of Kells famous

Elaborate incipit page for the Gospel of Matthew
Highly ornamented opening of Matthew with large initial letters filled with interlace and figures. Source: Trinity College Dublin

On folio 29r, the opening of Matthew begins with the Latin Liber generationis, ‘book of the genealogy’. Rather than simply enlarging the first letters, the artists filled them with interlace and figures until writing and picture became difficult to separate at first glance. The words still introduce Matthew’s Gospel, but the reader must follow the shapes through the composition to read them. The decorated page asks for careful looking as well as careful reading.

The best-known example appears on folio 34r, the Chi Rho page. Chi (Χ) and rho (Ρ) are the Greek letters traditionally used to abbreviate Christ’s name. Here they fill the page at Matthew 1:18, the point where the Gospel turns to Jesus’s birth. Figures of birds, other animals and people emerge from the strokes and interlacing. Moss describes an area measuring only about 80 by 45 millimetres containing three lions, four human figures, four snakes and thirteen birds. The painstaking miniature figures draw attention to the opening of the Gospel passage about Jesus’s birth.

The artists also illustrated religious subjects outside the decorated initials. Folio 7v shows the Virgin Mary with the infant Jesus, attended by angels, and is among the early surviving Western manuscript depictions of this scene. The traditional symbols of the evangelists appear as well: a human figure for Matthew, a lion for Mark, a calf or ox for Luke and an eagle for John. They linked four separate narratives to the Christian understanding of a shared message. Portraits of Matthew and John survive, whereas those of Mark and Luke do not. Some scholars suspect that additional illustrated leaves were lost, but the absence of the portraits cannot prove that they were ever completed.

The manuscript contains a great cross design with roundels, images of Christ and scenes relating to his temptation and arrest. Insular Gospel books had long combined Christian imagery with complex patterns, but the Book of Kells uses ornament with exceptional density and variety. Some of its animals seem playful; others are hard to identify. Set into the letters of Scripture, they encouraged sustained attention to a book whose physical form expressed the holiness and authority of its contents.

Why such a beautiful book contains mistakes

A Gospel manuscript of this scale would have been highly valuable to the community that commissioned it. A Gospel manuscript represented scriptural authority, and its decoration also displayed the learning and resources of its monastic community. It may have been shown on feast days or kept in a prominent church setting. Trinity’s suggestion that it had a primarily ceremonial purpose comes partly from the contrast between its sumptuous appearance and its many textual errors, which would have made repeated practical consultation less satisfactory.

We have no surviving instruction from the makers specifying when the book was to be displayed or read aloud. The ceremonial explanation is therefore a reasoned interpretation, not evidence for a particular liturgy. The splendid openings and the prominence given to Matthew 1:18 suggest that certain passages held special visual and religious importance. They cannot establish a complete cycle of readings or show that one ritual governed the decoration of every leaf.

Animals and geometric patterns run through the manuscript’s largest initials, whose shapes remain recognisable as letters. The decoration draws attention back to the Gospel passages it surrounds. For the manuscript’s early custodians, reverence for its sacred words could extend to the care and beauty invested in making the book itself.

The stolen Gospel and its journey to Dublin

The clearest early written evidence connecting a great Gospel book with Kells comes about two centuries after the manuscript was made. An entry in the Annals of Ulster, preserved in an edition of the Irish charters, reports the night-time theft of the great Gospel of Colum Cille from the stone church at Kells. The entry is normally assigned to 1006, or 1007 when the annal’s chronology is adjusted. After two months and twenty nights, the book was recovered beneath a sod, but its gold had been taken. Scholars generally identify this stolen volume with the surviving Book of Kells. The record does not identify it by the modern catalogue number MS 58, however, so that identification remains an inference.

The thieves may have been after the cumdach, a valuable shrine-like outer case or cover associated with a revered book. The annal’s account of missing gold therefore does not mean that someone stripped gold leaf from the painted pages. Removal of a precious covering may also explain the loss of leaves near the front and back, although this remains unproven. What the account does show is that a Gospel associated with Colum Cille was treated at Kells as a sacred relic with considerable material value.

Other evidence comes from writing added to the manuscript itself. During the later eleventh and twelfth centuries, members of the Kells community entered records of land transactions in unused spaces on several leaves. These Irish charters copied into the book show that the sacred volume also offered a respected place to preserve documents important to the monastery. A fifteenth-century poem written on folio 289v points to the book’s continued association with Colum Cille.

The Book of Kells remained in the monastery’s orbit until the political upheavals of the seventeenth century. According to Trinity’s account, Charles Lambert, the governor of Kells, sent the manuscript to Dublin for safekeeping around 1653, after the destruction connected with the rebellion of 1641 and subsequent wars. Henry Jones brought it to Trinity College in 1661. A book long treasured within a religious community thus came into the care of an academic library, where it would be preserved and studied under very different conditions.

Its physical appearance has continued to change. During nineteenth-century rebinding, the leaves were trimmed and their edges gilded; in 1953 the manuscript was rebound as four separate volumes. Trinity now displays it under controlled conditions and provides digital images that can be studied without repeatedly handling the vellum. The surviving object bears the effects of medieval use, missing leaves and later conservation alongside the work of its makers around AD 800.

The Book of Kells preserves evidence of how early medieval religious communities invested labour and resources in the Gospels. Its minute paintings, ruled parchment and copying mistakes reveal different parts of the same undertaking. We still cannot assign it confidently to Iona, Kells or a Pictish workshop, or reconstruct all the pages now missing. Its surviving parchment, colours, Latin text and later annotations allow us to examine a remarkable medieval book while tracing how successive custodians used and protected it.