The Bible Manuscript Hidden in Bookbindings for 800 Years

Somewhere inside the Megisti Lavra monastery on Mount Athos — one of the most storied Christian libraries in the world — a book was dying. It was a large, handsome codex of Paul’s letters written in ancient Greek majuscule script, possibly as early as the sixth century. At some point, its ink had faded badly enough that a monastic bookman reinked the entire manuscript by hand, tracing over the original letters to save them. The parchment pages had grown irregular, the flesh sides and hair sides of the animal skin reading differently under the light. It was, by the twelfth or thirteenth century, an uncomfortable thing to work with — and the monks of the Megisti Lavra, ever practical stewards of a large collection, quietly decided it was more valuable as raw material than as a book.

They stripped it apart. The pages that had carried Paul’s words for perhaps eight centuries were separated from one another, tucked inside the bindings of other manuscripts, as stiffening cartonnage and protective flyleaves. A codex that may have originally held over 400 folios — more than 800 pages — was swallowed piece by piece by the library around it, hidden in the guts of younger, healthier volumes, and effectively forgotten.

This is the story of Codex H (GA 015), and thanks to a remarkable new scholarly project led by Professors Garrick V. Allen (University of Glasgow) and Kimberley Fowler (University of Groningen), published in the Journal of Biblical Literature in 2025, that story is finally being told in full. HistoryMedieval put questions directly to Professor Allen about the project — and his answers illuminate not just the manuscript, but what it means to recover something thought permanently lost.

GA 015 Megisti Lavra sparagmata, 7v Tof the Megisti Lavra. Credit: The University of Glasgow

What Is Codex H — and Why Does It Matter?

Before we can appreciate how Codex H was lost, we need to understand what it actually is and why it has excited biblical scholars for well over a century.

Codex H is a majuscule manuscript — meaning it was written in large, separated capital letters, without the joined cursive script that developed later in the Byzantine period. Its text covers Paul’s letters in Greek, and its script, materials, and layout suggest production somewhere in the eastern Mediterranean, possibly Palestine or Syria, before it eventually arrived at the Megisti Lavra, probably in the tenth century. The date of its original production remains debated: the traditional view places it in the sixth century, while more recent palaeographic analysis by Elina Dobrynina suggests the eighth or ninth century is equally possible.

What makes Codex H exceptional is not simply its age but its connection to a remarkable scholarly tradition known as the Euthalian apparatus — a system of paratextual features that helped readers navigate the New Testament. These features include chapter lists, quotation markers, prologues to individual letters, biographical notes on Paul, and the unusual practice of arranging the text in “sense lines” or “thought units,” so that lines of continuous argument are visually grouped on the page. Codex H preserves the earliest known evidence of this apparatus in Greek, making it a uniquely important witness to how early Christian readers organised and approached Paul’s writing.

The chapter lists alone are worth pausing over. “These chapter lists are the main way that people divided Paul’s letters into sections up until the age of print, when our modern chapter and verses were devised,” Professor Allen explains. “Practically, the lists divide the letters into small sections and give each section a title.” The differences from modern readers’ experience of Paul can be striking: where the Letter to the Hebrews has thirteen chapters in contemporary Bibles, Codex H’s lists divide it into twenty-eight sections, each with its own heading. “Because the lists have titles for each section,” Allen says, “they are kind of like short summaries of the entire letter, giving readers a framework for what they are about to interpret.” That framework was, in effect, the interpretive lens through which Christians read Paul for nearly a millennium before the modern chapter and verse system displaced it entirely.

The manuscript also carries a colophon — a closing note — that may attribute the apparatus to a figure called Evagrius rather than the more commonly cited Euthalius. That first line is heavily abraded, and the reading is uncertain, but it hints that the origins of one of the most widespread paratextual systems in the New Testament tradition remain genuinely open questions.

A Living Manuscript: Readers, Correctors, and Annotators

One of the things that strikes any reader of Allen and Fowler’s research is how thoroughly Codex H was used. This was not a prestige object kept behind glass. It was a working book, handled, corrected, and marked up by generations of readers across several centuries.

When asked what the scribal annotations reveal about how people interacted with sacred texts during this period, Professor Allen’s answer conveys just how rich the evidence is. “The manuscript has around one hundred notes and other interventions by later readers that show how people used the manuscript over time, and they’re very diverse,” he says. “They include things like corrections to the text by multiple people, prayers, short poems, notes on Greek grammar, scribal practices, marks for reading the texts in the liturgy, the location of the manuscript in the library of the Megisti Lavra monastery, and ownership notes.”

The picture that emerges is of a manuscript that connected its readers not just to Paul’s theology but to the texture of their own daily lives. “Together,” Allen observes, “the notes show a group of active readers interested in the fine details of the text and in connecting the daily activities of their lives to the larger reality as they perceived it.” Someone found space to jot a prayer. Someone else noted a point of Greek grammar. A liturgical reader marked the passages to be read aloud. A librarian recorded which shelf the book stood on. These are not the traces of passive reverence but of genuine intellectual and spiritual engagement — a reminder that ancient manuscripts were not merely containers for text but participants in the lives of their communities.

Medieval Book Culture: The Logic of Dismemberment

To modern sensibilities, the idea of cutting apart a sacred text to use it as bookbinding material seems shocking. But within the practical world of medieval monastic book culture, it made perfect sense — and understanding that culture is essential to understanding what happened to Codex H.

The Megisti Lavra, founded in 963 CE, was not merely a place of prayer but a full production and conservation centre for manuscripts. It received volumes from Constantinople, Asia Minor, and other threatened monasteries across the Byzantine world. Over centuries, its shelves have been filled with manuscripts of varying age, condition, and utility. When a codex became too damaged to read comfortably, or when its text was redundant, the monks faced a practical question: what do you do with 400 large parchment folios?

Professor Allen frames the logic precisely. “Codex H enjoyed a very long working life, probably somewhere between 500 and 600 years,” he says. “It was conserved multiple times, including a point where it was entirely re-inked, which probably created the ghost text that we were able to uncover. So multiple communities worked to keep the book in working order for a long time.” When the book finally reached the end of that working life, the community’s response was not negligence but economy. “When it became clear that it had reached the end of the road, the community of the Megisti Lavra, which has one of the great libraries of the Greek world, used the material in the most economical way possible: to repair and conserve other books in their collection.“

And here is the central irony that runs through Codex H’s entire story: “Ironically, Codex H only exists still today in part because it was disbound and recycled,” Allen notes. “Because this practice was so widespread in the medieval world, one of the best places to discover unknown manuscripts today is in the bindings and flyleaves of already known copies.” The act of dismemberment that seems, at first glance, to be a destruction was actually a form of preservation — pages folded into bindings, shelved inside other volumes, waiting in the dark.

In 1218, a monk named Macarius was carrying out exactly this kind of conservation work at the Megisti Lavra. He was rebinding a tenth-century manuscript containing Gregory of Nyssa’s commentary on the Song of Songs. Looking for cartonnage material, he reached into a pile of loose leaves and selected two folios from Codex H — one from the end of Titus, one from a colophon mentioning the library of Caesarea. He placed them between the boards and the new pastedown, sewed everything together, shelved the result as the 34th book on the first shelf, noted the date in a large annotation, and added a formal curse against anyone who dared remove the book from the library. The two folios of Codex H would not be seen again for nearly five hundred years.

Ghost Text: Reading the Invisible Pages

One of the most extraordinary aspects of Allen and Fowler’s project is the recovery of what scholars call ink transfer — a phenomenon that is, in effect, ghost text left behind when the codex was still intact and closed.

When a manuscript is closed after fresh or caustic ink has been applied, the ink on one page can transfer a faint mirror image onto the facing page. Codex H was reinked at some point during its working life, and after that reinking, its pages were pressed together. The caustic ink continued to affect the facing parchment even as the book sat closed on a shelf. The result, centuries later, is that many pages of Codex H that no longer physically exist have left ghostly imprints on the pages that faced them — readable, at least partially, under multispectral imaging.

Scholars working in the late nineteenth and early twentieth centuries had already glimpsed this phenomenon using the photographic technologies of their time, but only for a selection of pages and with limited clarity. Allen and Fowler’s project, working with the Early Manuscripts Electronic Library (EMEL) in 2023, took the effort to a qualitatively different level. “We suspected that every page essentially had two texts,” Professor Allen explains: “the words it was intended to have and the text of the lost facing page (the ghost pages). By imaging each page 46 times in different exposure times and lighting configurations, we created a huge amount of new data that, when processed by EMEL’s data scientists, allowed us to create versions of the page that emphasised the imprint of the ghost text.”

The scale of what this revealed is striking. Working with the 41 known physical folios (82 pages), the team estimates that up to 62 additional pages of text may be at least partially recoverable through this method — bringing the total of potentially legible material to approximately 143 pages. For Professor Allen, the significance extends beyond the raw numbers. “Finding new material like this is important for so many reasons,” he says. “It helps us to understand how books were made and cared for, how people actually engaged their sacred texts, and what the text of the New Testament looked like in the sixth century, a period where we don’t have very many copies.” But the discovery also carries a more general lesson about what familiar objects can still yield. “It also shows that we have much to learn from old artefacts, even ones that are quite well known. When we take on new approaches, like the multispectral imaging and data processing we did in partnership with EMEL, it’s clear that old manuscripts have much new to tell us. In this case, Codex H gives us many unique snapshots into the lives of the real people who made and used it.”

The Long Dispersal: From Mount Athos to Seven Cities

After Macarius’s time, parts of Codex H continued to leave the Megisti Lavra over a period of roughly three centuries, carried away in the bindings of other manuscripts or removed directly as loose leaves. Today, its 41 known folios are held in eight institutions across seven locations: Paris, Torino, Kyiv, Moscow, Saint Petersburg, and the Megisti Lavra itself.

The stories of how each group of folios left Mount Athos are, almost without exception, stories of colonial-era acquisition — and frequently of dubious or outright dishonest dealing.

The two Torino folios arrived in Italy sometime between 1628 and 1629, likely as part of the collection of Gabriel Severos, an Orthodox ecclesiarch and bibliophile whose estate was purchased after his death by Charles Emmanuel I, Duke of Savoy. Twelve folios now in Paris arrived in the late 1640s as part of the Coislin collection, assembled for the French nobleman Pierre Séguier through the agency of Athanasius Rhetor, a Cypriot-born priest who secured access to Athos with Ottoman diplomatic assistance. Dositheos II Notaras, Patriarch of Jerusalem, later described Athanasius’s methods with striking frankness: he visited monasteries in disguise, selected manuscripts, and paid for them by weight in silver — buying books as if they were scrap metal, not knowledge.

Two Moscow folios were taken by Arseny Suchanov, a Russian monk and diplomat who returned from Athos in 1655 with over 500 manuscripts. Three Saint Petersburg folios were removed in the nineteenth century — two by Piotr Dubrovsky, a Russian Embassy official who acquired them in Paris during the revolutionary 1790s, and one by Bishop Porphyrius Uspensky, whose own correspondence was met with letters from Athonite monks demanding the return of stolen items. Ten further folios, now held in Paris as Supplement grec 1074, were taken by the French scholar Emmanuel Miller and the archaeologist Louis Duchesne during expeditions to Athos in the 1860s and 1870s under the auspices of Napoleon III’s Académie des Inscriptions et Belles-lettres. Miller’s published travelogue characterised the monks as ignorant and barbarous — claims that Allen and Fowler demonstrate are entirely unsubstantiated, serving primarily to justify extraction.

Colonial Aphasia and the Ethics of Manuscript History

Allen and Fowler engage directly with what scholar Jennifer Wright Knust calls colonial aphasia — the rhetorical pattern by which Western collectors framed their acquisition of manuscripts as acts of salvation, simultaneously erasing the agency and knowledge of the communities from whom those manuscripts were taken. The monks of the Megisti Lavra, who had preserved Codex H for a millennium, do not appear in the traditional history of the manuscript at all. The acquirers’ accounts — partial, self-serving, and in several cases deliberately silent about the circumstances of acquisition — are all that remain.

This is not merely an ethical footnote. Allen and Fowler argue that understanding how we got the manuscripts is a necessary precondition for understanding the manuscripts themselves. The shape of primary evidence for New Testament scholarship is not neutral: it reflects centuries of unequal power relationships, national rivalries, and colonial attitudes that moved material from Orthodox monasteries to European state libraries. Codex H, dispersed across seven cities by the acquisitive energies of bibliophiles, diplomats, and scholar-pilgrims, is a microcosm of that broader story.

The researchers’ response is restorative in intent as well as method. Following Allen’s visits to the Megisti Lavra in 2023, the team has invited Fr. Athanasius, Secretary of the Synaxis and keeper of the Library, to write a foreword to their forthcoming edition — formally acknowledging the monastery’s role as guardian of the manuscript for over a thousand years.

What Comes Next

Allen and Fowler, working with collaborators Emanuele Scieri and Maxim Venetskov, are preparing a new digital edition of Codex H that will reconstruct the manuscript as it existed before its disassembly. This will draw on the multispectral imaging completed in 2023, the ghost text recoveries, and detailed codicological analysis of surviving folios, including their quire signatures, text layout, and material features.

The reconstruction is genuinely complex. The original manuscript may have contained over 800 pages and probably included Acts and the Catholic Epistles in addition to Paul’s letters, though this remains a working hypothesis. Its unusual layout and early, consistent use of diacritical marks make it a remarkable witness to ancient reading practice. And the colophon, partially legible under infrared imaging, continues to pose questions about the identity of the scholar behind the Euthalian apparatus.

What Codex H ultimately offers is something rare: a case study complete enough to follow a single object from its probable production in the eastern Mediterranean, through its arrival at a major Byzantine monastery, through its practical dismemberment and dispersal across the collections of Europe and Russia, and back again — at least in digital form — toward a coherent whole. A book that survived by being hidden, and that continues to yield new knowledge precisely because scholars were willing to look more closely at what everyone assumed had already been fully seen.