On 14 April 1705, in a house in Amsterdam where a professor of the gymnasium had died, the sale catalogue of his library came to the hammer. On the third page, as item eighteen, there was a small book of parchment containing the Epistles of Saint Paul, in Greek, with a Latin version written between the lines. It was bought for a hundred and forty-two guilders by a Leipzig theologian travelling through Holland with a pocketful of money and an eye for treasures. The book took the name of that theologian, and for three hundred years it was known accordingly as the Codex Boernerianus.
It is worth being precise at the outset about what this object is not, because it has been confused more thoroughly than almost anything else in the history of Latin manuscripts. It is not a Gospel book. It contains no Gospels at all. It is not from Wessobrunn, and it is not in Oldenburg. It is a manuscript of the Pauline Epistles, from Romans to Philemon with Hebrews missing, written at St Gall in the second half of the ninth century and now kept in Dresden, in the Sächsische Landesbibliothek, under the shelfmark Mscr. Dresd. A. 145.b. It carries the siglum G and the Gregory-Aland number 012. And its Latin is not a translation, in the sense that most catalogues mean by a translation. It is an apparatus, written word by word above the Greek, and it is that apparatus which has turned a book long dismissed as a backward and marginal curiosity into the centre of a live argument about how early the recoverable text of the Greek New Testament runs.
The argument is being run digitally, by the TextLab of the Sächsische Landesbibliothek in Dresden together with the chair of New Testament and Greek-Roman culture at the Ludwig-Maximilians-Universität in Munich, and it is the argument rather than the ornament that makes the book worth an essay. It is also a close parallel for anyone who has spent time with the masorah: a scribe writing a commentary on a base text, aligned to it word by word, in the same hand as the base text itself. The Boernerianus does to Greek what the Masoretes did to Hebrew, about four centuries earlier.
A Book Bought in Amsterdam in 1705
The first known owner of the manuscript was Paulus Junius, rector of the gymnasium at Leiden, from whom it was borrowed for collation by the classical scholar J. F. Gronovius. When Junius died in 1670 the codex passed to Peter Franz, who taught at the Amsterdam gymnasium and had it collated in 1690 by A. Bynæus. After Franz’s death it came into the sale of 1705 that Börner bought it from, and the catalogue that records the transaction lists it at three pages in, as number eighteen. The buyer was Christian Friedrich Börner, afterwards professor of theology at Leipzig, and it is from him that the manuscript has taken the name by which it is still filed.
Börner did not simply own the book. He worked on it. Within a few years he had sent it to Ludolph Küster in Holland, who collated it and reported its variants for a new edition of Mill’s New Testament, published at Amsterdam and Leipzig in 1710. Richard Bentley, preparing his own edition, then asked to borrow it, and had it copied – the copy survives in Trinity College, Cambridge, as B. 17.2 – and only surrendered the original after five years and the intercession of a bishop.
Wetstein, preparing his Amsterdam edition of 1751 and 1752, used transcripts of the Gronovius and Bynæus collations and gave the manuscript the siglum G – the siglum everyone still uses. When Börner’s library came up for auction in 1754 the family bought the codex back for seventy-three thalers. In 1777 it was offered to the Electoral, and shortly to be Royal, Public Library at Dresden; the librarians reported on 9 January that it could hardly fail to be acquired, given its age and its generally recognised value, and estimated it at not less than a hundred ducats. It was bought in March 1778 for two hundred and fifty thalers in silver, the price negotiated in Leipzig with Börner’s grandson, Dr Wolle. It has been in Dresden ever since.
Before the seventeenth century there is a silence, and it is a silence with a shape. The library of St Gall was plundered more than once in the fifteenth century – in 1414, during the Council of Constance, and again in 1431, during the synod of Basel – and Heinrich Zimmer, who did more than anyone to establish what the manuscript is, plausibly conjectured that it went out with the loans made to the fathers at Constance and never came back. That is a conjecture, and it should remain one; the codex in Dresden is separated from the abbey that most probably produced it by an unknown interval of four or five hundred years, and its early history has to be reconstructed from what happens to it once it surfaces in Leiden.
The physical description is correspondingly firm. The codex consists of a hundred and eleven parchment leaves, a little over twenty-five centimetres tall and about nineteen wide, one side of each leaf smooth and the other rough, with two smooth and two rough sides falling together in the regular quires. The first leaf and the last eleven carry no numbers; the ninety-nine leaves between them are numbered 1 to 99, and it is on those ninety-nine that the text of the Epistles stands. Romans begins on the recto of the first numbered leaf, and the letter to Philemon breaks off in its closing verses on the verso of leaf 99. In the last three quires there are traces of excised leaves, which is where the later addition to the volume came from. The Dresden library describes the book as measuring twenty-five by eighteen centimetres. The binding is old and much damaged: wooden boards under yellowish-grey wild leather, worm-eaten, with the remains of three clasps and the impression of a fitting where a chain was probably once attached.
The Latin That Is Not a Translation
Look at a page of the manuscript and the first thing you see is two layers of text. The Greek runs as a continuous block; above it, standing on its own line and aligned to the Greek beneath, stands Latin. The Dresden project describes the arrangement exactly: the Latin stands on a separate line above almost every word of the continuous Greek text, with one and sometimes up to three equivalents.
It is tempting to call the upper column a translation, and for a long time that is what it was called. The 1791 edition by Christian Friedrich Matthäi, the Wittenberg professor who produced the first printed text of the codex, titled his book with a description that now sounds like an artefact of a moment: the codex carried a Latin version “veteri vulgo antehieronymiana”, the old version commonly called pre-Hieronymian. Modern scholarship would not accept that. Ludwig Ziegler, whose study of the Latin Pauline fragments appeared in the 1870s, put the better view: the interlinear column reproduces none of the known pre-Hieronymian versions exactly, and does not claim to. It is nothing more than a means of understanding the text beneath – a simple interlinear version rendering each individual word of the original by an expression as nearly equivalent as possible.
That is a gloss, and it behaves like one. The scribe could use one or several of the translations then in circulation, and contaminate the older renderings with the Vulgate, and the evidence is everywhere on the page. At a good number of points the Greek reading of the base text descends from a scribal slip, and the slip has carried over into the Latin above it, so that the Latin silently corrects the Greek it is supposed to be explaining. Elsewhere the same Greek word is given two or three Latin equivalents side by side, one of which is usually the Vulgate’s. The result is not a text that can be collated like a text. It is a text that can only be read, and read slowly, by someone learning how the Greek works.
Two marginal notes make the working method explicit. On the verso of leaf 23 the scribe has written that something “deest in graeco”, is missing in the Greek; on leaf 32, that something “non est in latino interpretatum”, is not rendered in the Latin. Both point to the same conclusion: more than one exemplar was in play, one carrying the Greek text and at least one carrying a Latin translation with commentary upon it. It has therefore been suggested that the codex was used for the study of Greek in a monastic school, with the Latin column as a crib.
The hand itself tells the same story from a different direction. The Greek is written in a small uncial. The Latin is written in Anglo-Saxon minuscule, with words separated and a point very frequently set between them – though the editors are agreed that most of those points are not punctuation at all but word-division marks. The large Greek initials and the name of the letter written directly above the text are picked out in cheap colours; there is not one artistically executed initial in the book and no painting of any kind. The impression is of a working book, made to be used.
Around the text is a fringe of notation. In the margins are chapter numbers, at first in both Arabic and Roman numerals and later in Roman only, with verse numbers here and there; a quantity of signs, names and short remarks, most of them written at about the same time as the main text; a Latin hexameter at the top of leaf 25, and a Latin pentameter at the upper left of leaf 82. Among the personal names is one written in Greek and transliterated, “goddiscalcos”, with a note that it is against or concerning Godiscalcus – an Irish name, and one of the threads pulled to identify the scribe. The most striking item is on leaf 23, below the Greek and Latin text and in a smaller and later script: a short poem in Old Irish, published by Franz Delitzsch at the end of the nineteenth century and then printed in full in the second volume of the Thesaurus Palaeohibernicus. Nestle’s Introduction prints the opening stanza as it was then rendered into English: “To Rome to come, to Rome to come, much of trouble, little of profit, the thing thou seekest here, if thou bring not with thee, thou findest not.” It is a complaint about a journey to Rome, and whether its author is the copyist of the Epistles or a later reader is something the manuscript does not settle. It is this poem, together with the script, that has led scholars to infer that the manuscript was written by an Irish monk, probably at St Gall. That is an inference, and a good one, but it is an inference: the poem is evidence for a scribe of Irish origin, and the connection with St Gall is a further step, not a statement made by the manuscript itself.
What Was Missing, and Why It Matters
Open the codex at any point and there is a reasonable chance of finding a gap. The Vetus Latina’s catalogue records that the following portions are missing: Romans 1:1-5, 2:16-25 and 16:25-27; 1 Corinthians 3:8-16 and 6:7-14; Colossians 1:1-8; and Philemon 21-25. Reichardt’s 1909 description lists the same losses differently, because it identifies them by the Greek words at which the text breaks off rather than by Latin verse numbers, and among its list are the end of Colossians and 2 Timothy 2:12-13. The two lists are not in conflict. They are two ways of measuring the same holes, in a manuscript whose Greek and Latin carry different verse divisions – and the disagreement is a small demonstration of why a manuscript like this needs more than a table of contents to describe it.
What matters more is the shape of the damage. These are not losses inflicted by time. They are spaces left blank on purpose. The copyist came to a place where his exemplar lacked text, and instead of closing the gap, or writing something in it, or leaving a mark, he left the room clear. The Dresden project makes the point directly: where a large part of the first five verses of Romans and the last three are absent, this, they write, is not an oversight on the part of the copyist, for he left extra space in order to add the supposedly missing text at a later date. Someone was expected to come back.
One of those gaps is among the most consequential in the history of the Latin New Testament. The received text of Romans ends with a doxology – a passage in three clauses, praising the God who alone is wise and glorious, keeping and confirming the saints, which in the critical Greek tradition is numbered Romans 16:25-27. The Boernerianus does not have it in that place. Instead, at the close of chapter 14, the scribe leaves five and a half lines free. The space is exactly the right size. The received doxology has been reserved at the point where an older arrangement of the letter put it, and it was never filled.
This one gap has done more to shape the manuscript’s reputation than anything else in it. The Codex Augiensis, a ninth-century bilingual of the same family, also lacks the doxology – although it lacks it differently, leaving its blank space for the Greek at the end of chapter 16 and supplying the Latin translation of the doxology there. And a Bobbio manuscript in the Ambrosiana and a manuscript at Monza, both showing kinship with the Boernerianus, also lack it. Zimmer’s demonstration of 1881 clinched the relationship between F and G, as the next section shows. The oldest recoverable arrangement of the letter, at least for this family of witnesses, had no doxology at the end of it, and the doxology now printed at Romans 16:25-27 belongs to a later history of the text.
The rest of the book has one further surprise, which is at the end. After a blank space reserved for the conclusion of the letter to Philemon comes a heading in Greek and Latin – “ad laudicenses incipit epistola”, the letter to the Laodiceans begins here – and the heading of a letter that is not in the volume at all, since the apocryphal Epistle to the Laodiceans was never copied here. What follows instead is a Latin explanation of Matthew, running on to about Matthew 5:22 before ending, at the very bottom of the last page of the codex, with the words “per validam iudicii sententiam in gehennae ignem mittetur, qui hic boni operis fructum fecisse non invenitur”. Houghton identifies the commentary as that of Otfrid of Weissenburg. In the upper left of the same page, in eighteen lines of Greek with a Latin interlinear version, is a fragment of a treatise on the spiritual law by a certain Marcus monachus – possibly, if the handwriting is any guide, by the same hand that copied the Epistles. Reichardt’s account of how this happened makes the book whole: the commentary was added later, by a different hand, to a volume whose leaves had been cut away, and the newcomer used up what parchment remained after the excisions, disguising his addition by placing it at the beginning and at the end of the codex where a reader would take it for original. The manuscript, in other words, was in use for a second purpose, by a second reader, several centuries after it was written; and it was still being repaired, still being expected to be supplied with the passages its first scribe had left open.
Two Centuries of Dismissal
The book was noticed, and then it was not. The first full description was printed by Börner as an appendix to Le Long’s Bibliotheca sacra in 1709; Küster, collating it for the edition of Mill that followed, had already observed that the script is Insular and the ninth century the only date that fits; and Matthäi put it into print in 1791, with plates. Church’s study of 1842 bore the title The Church in the Middle Centuries: an attempt to ascertain the age and writer of the celebrated Codex Boernerianus, and Scrivener gave the manuscript a paragraph in the third edition of his Plain Introduction in 1883. Nestle’s Introduction, the standard handbook of the Greek New Testament, got through the manuscript in a single paragraph. G, it records, is of the ninth century, so called from Professor C. F. Boerner of Leipzig, who purchased it in 1705, and is now in Dresden; it is a Greek-Latin manuscript, the Latin being interlinear. Then the observation that was to stand for a century: this manuscript contains some interesting Irish verses. At the end of Philemon there stands the title ad laudicenses, but the Epistle that should have followed has been lost.
None of this made the manuscript matter to the text of the Greek New Testament, and the reason is instructive. Lachmann, whose work on the Greek text in the 1840s did more than anyone else’s to set the terms for the discipline, praised the Greek of the Boernerianus in 1842 as very valuable, and Gregory called it good. What they could not use was the Latin. In an age of eclectic editing, in which a manuscript earned its place by supplying readings that could be weighed against other witnesses and the Fathers, an interlinear gloss written by a monk working for his own pupils was not the kind of evidence that could be weighed. It was too idiosyncratic, too contaminated with the Vulgate, and too hard to separate into the components a critic would need. The codex was not so much rejected as set aside – and then set aside for a very long time.
The instrument that changed that was a facsimile. In 1909 Alexander Reichardt edited, in Leipzig with Karl W. Hiersemann, a monochrome reproduction of the whole book under the title Der Codex Boernerianus; der Briefe des Apostels Paulus (Msc. Dresd. A 145b) in Lichtdruck nachgebildet, with a descriptive preface that remains the most valuable single document in the manuscript’s history. It records the physical state of the volume leaf by leaf, the excised leaves, the binding and its clasps, the marginal notes, the Old Irish poem on leaf 23, the coloured initials and the absence of any decorated ones. It transcribes the ownership history from Junius and Franz through Börner and Bentley to the Dresden purchase of 1778. And it sets out, with the arguments and the counter-arguments, the whole dispute about the relationship of the Boernerianus to the Augiensis and the Claromontanus, at a point when that dispute was still being fought in print.
That facsimile was made before the accident of 1945, when the manuscript suffered serious water damage. The Vetus Latina’s register records it as a fact of the object’s history, without describing what was lost. The precise extent of what was lost or obscured is best judged from the plates rather than from the volume as it stands, and the 1909 reproduction is the reason that the question can be asked at all. That is the ordinary condition of codicology: the facsimile is not a substitute for the object, but it is often the only witness to what the object used to be, and it preserves the state of the codex at a moment when it was still intact enough to be collated in its entirety.
X and Z: Recovering a Lost Ancestor
Three bilingual manuscripts of the Pauline Epistles survive from the first millennium. They are the Codex Claromontanus, called D or 06 in the Gregory-Aland sequence and copied in the sixth century; the Codex Augiensis, called F or 010 and copied in the ninth; and the Codex Boernerianus, G or 012, which is the subject of this essay. The Dresden project calls the survival of all three a circumstance that is a remarkable exception in the New Testament manuscript tradition: a tiny number of books preserve the Pauline text with the Greek and its Latin side by side, early enough to be worth asking what they were copied from.
It was also already recognised at the end of the nineteenth century that the Augiensis and the Boernerianus must have had a common template, conventionally called X. The hypothesis was set out with some rigour by P. Corßen, whose two studies of 1887 and 1889 bore a title that states the argument in Latin: he examined the three manuscripts against each other and sought to reduce them to a common origin. Corßen argued that F and G were copied from a lost exemplar X, and that X itself went back with the Claromontanus to a common original.
He was not allowed to have it his own way. Heinrich Zimmer, who had done for the marginalia and the script what the nineteenth century could do, refused to believe in X and maintained instead that the Augiensis is a copy of the Boernerianus – a position he set out in 1887 in an article entitled Der Codex Augiensis eine Abschrift des Boernerianus, and which the twentieth century largely accepted, against Corßen, though Nestle dissented. J. B. Lightfoot took a middle course in the Journal of Philology, holding that F and G must go back, directly or indirectly, to the same original text. F. J. A. Hort thought F a copy of G, and the New Testament in the Original Greek that Westcott and Hort published in 1896 declared it practically certain that F contains a copy of G’s Greek text. Scrivener, who had himself edited the Augiensis, and Bentley before him thought that each of the two was copied independently from an older and now lost codex; so did Tregelles, and Tischendorf with greater confidence. The disagreement was never about the evidence, but about how much weight a shared error should carry.
Zimmer’s own demonstration has the peculiar clarity of a thing that can simply be looked at. At 1 Corinthians 6:6 the scribe of the Boernerianus carries on after verse 6 with words that belong to verse 15, and only then notices that something has fallen out of his exemplar – exactly what happens when a copyist is working from a defective text and does not realise it until he has written too much. In the Augiensis, the blank space left for the Greek at the end of Romans 16 could hardly have been overlooked by a copyist who had the Boernerianus or anything like it in front of him. The relationship was therefore not one of descent but of common ancestry.
Modern description of the Augiensis confirms how different the two books are in kind, and how alike in their source. The Augiensis is a parallel text: Greek in the inner column of each page, Latin in the outer, the Greek taking up rather more than two thirds of the width, the Latin corresponding largely to the Vulgate. The Boernerianus is an interlinear gloss, with Latin above the Greek and no parallel columns at all. The Augiensis lacks Romans 1:1-3:19, and for the Epistle to the Hebrews it has no Greek whatsoever, the Latin running across both columns. And yet the Greek of the Augiensis derives, as Houghton’s catalogue puts it, from the same source as the Greek of the Boernerianus. Two books of completely different type, sharing an ancestor: which is what X was invented to describe. The readings that carry the shared ancestry are the kind that survive in an ancestor’s slips: Nestle singles out sicut cancer ut serpat, “like a cancer that creeps”, at 2 Timothy 2:17, and eramus autem servientes, “we were indeed slaves”, at Galatians 4:3.
The third member of the diagram is the Claromontanus, and the step from X to the older manuscript carries the stakes. The Dresden project’s own account is that the Claromontanus is said to have been created on the basis of an original Z, shared with X. So the diagram runs backwards through time: D was copied from Z; Z and X descend from a common ancestor; X is the template of F and G; and the Boernerianus and the Augiensis stand at the far end of the chain, in the ninth century, in an Irish and a Reichenau hand. The Claromontanus is sixth century, and therefore the recoverable history of the Pauline text can be pushed back behind it, towards the third century – which is the point the project makes when it says that the writing provides an extremely interesting insight into textual history up to the third century.
X and Z are working hypotheses in the sense that every stemma is, and the comparison on which they rest is being redone from the beginning with new tools.
Those tools are the reason the question has reopened. The joint project of the SLUB TextLab and the chair of New Testament and Greek-Roman culture at the Ludwig-Maximilians-Universität in Munich begins, as its own account describes, by typing the text of each manuscript out word for word as it appears on the page, and by recording every relevant graphic feature – line breaks, conspicuously different letters, corrections – in TEI-XML, the mark-up language used for scholarly editions. The collation can then be done both by hand and by computer. The interesting part is the emphasis the project places on the graphic characteristics and the peculiar spellings, which it describes as extraordinarily important clues to the ancestry of the manuscript texts, and which allow conclusions to be drawn as to who copied whom. That is the inversion of the nineteenth-century story: Zimmer needed a single slip at 1 Corinthians 6:6 to prove a relationship, and the digital collation is built to find thousands of such slips and to ask, for every one, which of the three books came first. The demonstration in hand is the Epistle to the Galatians.
It is worth being clear about what the Boernerianus contributes, because it is not the obvious thing. Its Latin was never going to be a witness in the ordinary sense; it is a gloss, contaminated with the Vulgate and shaped by one monk’s attempt to make Greek legible. Its Greek, by contrast, is a text of real quality, praised by Lachmann in 1842 and called good by Gregory, and its lacunae and scribal slips are not defects but data. Two hundred years of dismissal turned out to rest on a misreading of what kind of object this was: the manuscript was not put aside because it was corrupt, but because the discipline could not yet use the only part of it that mattered.
The Gospel Manuscript That Is Not This One
It is worth returning to the correction with which this essay began, and to state it once more without qualification. The Codex Boernerianus is a Pauline manuscript. It contains the thirteen Epistles of Paul, with the Epistle to the Hebrews omitted, and it contains no Gospels whatever. It was written at St Gall, and it is in Dresden. The description of a Gospel manuscript in a Western order, produced in Germany and kept at Oldenburg, does not correspond to it, and the shelfmark that is sometimes attached to that description cannot be traced in the registers of Old Latin Gospel witnesses or in the CSNTM catalogue. Oldenburg’s medieval manuscripts are filed under a different series of signatures, and a manuscript from that library should be quoted from its own catalogue before it is described at all.
There is, however, a ninth-century Greek-Latin interlinear Gospel diglot with an Irish hand, and it is one of the closest companions of the Boernerianus in the world. It is the Codex Sangallensis 48, kept at St Gall in the Stiftsbibliothek and given the siglum Cod. Sang. 48, the Gregory-Aland number 037, and the Vetus Latina number 27, with the siglum delta. The Vetus Latina’s register of Old Latin Gospel manuscripts describes it in a single line: four Gospels, Matthew, Mark, Luke and John, copied at St Gall. The digitisation catalogue at e-codices in Freiburg adds that it was written by Irish monks, probably in the cloister of Bobbio about 850. Houghton’s catalogue adds the codicological detail: a hundred and ninety-nine folios, a little over twenty-two by eighteen centimetres, a single column of nineteen to twenty-seven lines, Irish minuscule, black ink with rubrication in red and yellow; John 19:17-35 missing. The registers differ slightly on the date, one giving 860/70 and the other 960/970, so the safe description is the second half of the ninth century.
Two of its features are directly relevant. The first is that Houghton’s catalogue calls the Codex Sangallensis 48 a sister manuscript of the Boernerianus: the two books stand in the same relationship as the Augiensis and the Boernerianus stand to the Claromontanus. The second is that the Latin of the Saint Gall Gospels is described as based on a form of the Vulgate close to that of the Egerton Gospels, conformed to the grammar of the Greek, and as preserving little Old Latin evidence. So even the phrase “the Greek-Latin Gospel diglot with an Old Latin text” needs qualifying. The description is not a passport but an approximation, and approximations are exactly what travel badly.
The order of the books is a case in point. The Vetus Latina describes the standard sequence in Old Latin Gospel books as Matthew, John, Luke and Mark, and the Codex Sangallensis 48 is in the order Matthew, Mark, Luke, John. Neither is the sequence that the phrase “the Western order” is supposed to denote, and no manuscript should be described as uniquely or pre-eminently Western in its order without a specific citation. The same caution applies to a group of striking Old Latin readings often attached to a Gospel manuscript of this kind: the passage usually called the weary Christ at John 4, the reading usually called the ruler of the Jews, the Gospel of the Hebrews at John 4:37, and the Five Hundred and Eighteen Years at John 5:1-2. These are readings of the Latin text of John, to be checked in the Vetus Latina Iohannes apparatus and in Houghton’s dedicated study of the Latin text of John in the Saint Gall bilingual Gospels. Whether any one of them stands in the Codex Sangallensis 48 is a separate question, and none of them has anything to do with a Pauline codex in Dresden.
The methodological lesson is the reason the collision is worth an essay rather than a footnote. Titles travel; shelfmarks do not. A name attached to a collector’s family, or to a manuscript that was once famous, becomes a handle, and handles are passed from hand to hand in the secondary literature with very little checking. A shelfmark, by contrast, is dry, local and attached to one physical object, and so it is exactly the sort of thing that gets abbreviated, misremembered or dropped. A reader who trusts the title more than the shelfmark will find that the book under the title is not the book the title has been describing for a century. The cure is dull and reliable: open the catalogue, find the shelfmark, and describe what is on the shelf.
References
- Codex Boernerianus – Joint Research on the Pauline Epistles — Sächsische Landesbibliothek Dresden, SLUB TextLab research project page, accessed 2 November 2026
- Codex Boernerianus – Gemeinsame Forschung an den Paulusbriefen — Sächsische Landesbibliothek Dresden, SLUB TextLab research project page, German edition, accessed 2 November 2026
- Der Codex Boernerianus (Msc. Dresd. A 145b), full digitisation — Sächsische Landesbibliothek Dresden, SLUB digital collections, accessed 2 November 2026
- Manuscripta mediaevalia: die mittelalterlichen Handschriften der SLUB — Sächsische Landesbibliothek Dresden, manuscripts portal, accessed 2 November 2026
- Der Codex Boernerianus; der Briefe des Apostels Paulus (Msc. Dresd. A 145b) in Lichtdruck nachgebildet — Alexander Reichardt, ed., Leipzig: Karl W. Hiersemann, 1909, Internet Archive facsimile, accessed 2 November 2026
- Der Codex Boernerianus, full-text transcription of the descriptive preface to the 1909 facsimile — Alexander Reichardt, ed., Internet Archive, 1909, accessed 2 November 2026
- Manuscripts of the Old Latin New Testament, register including VL 77 Codex Boernerianus — Vetus Latina Institute, University of Birmingham, accessed 2 November 2026
- Old Latin Gospel manuscripts, register including VL 27 Codex Sangallensis 48 — Vetus Latina Institute, University of Birmingham, accessed 2 November 2026
- Vetus Latina Iohannes: manuscripts — Vetus Latina Institute, University of Birmingham, accessed 2 November 2026
- Concordance of sigla for Old Latin New Testament manuscripts — Vetus Latina Institute, University of Birmingham, accessed 2 November 2026
- Manuscripts of the Old Latin Bible, master register — Vetus Latina Institute, University of Birmingham, accessed 2 November 2026
- The Latin New Testament: A Guide to its History, Texts, and Manuscripts — H. A. G. Houghton, Oxford University Press, 2023, open-access copy at the University of Birmingham PURE repository, accessed 2 November 2026
- The Latin New Testament: Appendix, catalogue of Latin New Testament manuscripts — H. A. G. Houghton, Oxford University Press, 2023, publisher landing page, accessed 2 November 2026
- The Latin New Testament, chapter 8: The Text of the Early Latin New Testament — H. A. G. Houghton, Oxford University Press, 2023, publisher landing page, accessed 2 November 2026
- Codex Boernerianus (GA 012), manuscript record — Center for the Study of New Testament Manuscripts, accessed 2 November 2026
- Codex Sangallensis (GA 037), manuscript record and digitised images — Center for the Study of New Testament Manuscripts, accessed 2 November 2026
- Codex Delta (Cod. Sang. 48): the Gospels in Greek with an interlinear translation into Latin — e-codices, Universitätsbibliothek Freiburg, accessed 2 November 2026
- The Latin Text of John in the Saint Gall Bilingual Gospels (Codex Sangallensis 48) — H. A. G. Houghton, in At One Remove: The Text of the New Testament in Early Translations and Quotations, Gorgias Press, 2021, accessed 2 November 2026
- Introduction to the Textual Criticism of the Greek New Testament — Eberhard Nestle, with the classic description of the Codex Boernerianus, Internet Archive full text, accessed 2 November 2026
- Codex Boernerianus, facsimile record of the 1909 Hiersemann edition — facsimiles.com, accessed 2 November 2026