The Saint Christopher woodcut dated 1423

Written by Ed Potten, formerly Keeper of Printed Books at the John Rylands Library and Head of Rare Books at Cambridge University Library. Ed is currently the Principal Consultant on the Werck der Bücher project in collaboration with the University of Manchester, Johannes Gutenberg University Mainz and the University of Erlangen, and on the Incunabula Cataloguing Project at the John Rylands Library. He has published widely on book and library history, with a particular focus on the fifteenth century.

At the start of the fifteenth century, woodcutters began to produce blocks intended for printing. Images, then later images and words, were cut in reverse into wood, the blocks were inked, a piece of paper or parchment laid over them, then friction applied to produce an impression.

There are around 70 surviving woodcut impressions known that can securely be dated to before 1440 and based on stylistic appearance these were produced by just a few workshops. From the late 1430s onwards, however, the number of single leaf woodcuts increased significantly, and designs became more complex and nuanced, often with text accompanying images.

The John Rylands’ St Christopher woodcut is perhaps the most famous impression of the fifteenth century. It is the only surviving example printed from this block and was discovered pasted into a manuscript in the famous library of the Swabian monastery of Buxheim in 1769. In 1803 it was acquired by George John (1758–1834), 2^nd Earl Spencer, whose interest was piqued by the fact that it is one of only a handful of early impressions to bear a date, in this case 1423 (“Millesimo cccc° xx° tertio”). Because of the presence of this date, the St Christopher was lauded as the earliest dated piece of Western European printing.

There was, however, disquiet from the outset. There is a fundamental problem with the St Christopher: it looks nothing like the 70 woodcuts we can securely date to the 1420s and 1430s. Stylistically, it sits much more comfortably with a series of other prints, images in stained glass, and other iconographic representations from the 1440s or 1450s. Furthermore, in every other case where a fifteenth-century print bears an early date, we can show that it does not relate to the date of production. The date of 1418 on a print of Madonna with Female Saints, for example, is thought to be the date not of the block, but of the drawing from which the block was copied; a date of 1446 which appears in a Saint Nicholas of Tolentino is likely to relate to the year of the saint’s canonization rather than the date of the woodcut.

Recent attempts to contextualise the print, and the manuscript in which it is pasted, have failed to resolve the issue of its date. In 2014, however, with a combination of the high-tech (multispectral imaging undertaken by the Digitisation Unit at the John Rylands Library) and the low-tech (an old-fashioned rubbing taken with a soft graphite pencil) a new piece of evidence was unearthed – a watermark in the paper on which the St Christopher is printed.

Watermarks are now widely used to date books, prints, drawings, and other media. Over many years, scholars have built up databases of watermarks found on papers which bear hand-written dates – account books, letters, dated manuscripts – and this data can reliably be used to date other examples of paper bearing the same or similar watermarks. So, if we can find a specific watermark or a characteristic grouping in the databases which matches that found in the paper of the St Christopher, we can infer that the print was probably produced within two or three years of those dated examples.

There are no watermarks within the databases on papers used in the early 1420s that share the St Christopher’s ‘Ochsenkopf’ watermark’s combination of size, narrow muzzle, inward-curving horns, a featureless face and a single stemmed cross surmounted with a six-pointed star. There are, however, a series of watermarks which share precisely these characteristics. Evidence of the use of these papers dates them to between 1428 and 1435. The papers were manufactured in Ravensburg, probably in one of two mills owned by Hans Krieger and Haintz Wolfertshofer, which were founded c. 1430, and the papers were extensively used within a geographical area centred around Ulm.

This new watermark evidence does not decisively settle the issue of date. However, when placed alongside stylistic and other evidence, a strong case can be made that both impression and woodblock are much more likely in date to the late-1430s. The art historical evidence for a later date is strong, and the very presence of xylographic text on the St Christopher reinforces this: none of the 70 or so woodcuts we can reliably group together as the earliest extant examples contain text. The characteristics of the watermark link it to a group of papers which were in use in Swabia between 1428 and 1435, and which are not found in papers prior to 1428. Finally, stylistically, the St Christopher’s origins can likely be traced to stained glass, panel paintings, and altarpieces from Ulm dated to the 1430s, over a decade after the xylographic date. We cannot know today why the cutter dated the block to 1423, but the paper and stylistic evidence suggests that it was likely produced in the mid- to late-1430s.

None of this, however, detracts from the print itself; irrespective of its date, it remains one of the most remarkable and beautiful fifteenth-century impressions to survive.

This blog post is an abstract from a longer article, the full reference of which is:

• Edward Potten, ‘A Succession of Uncertainties: Dating the Buxheim Saint Christopher’, in John Goldfinch, Takako Kato, and Satoko Tokunaga (eds.), Production and Provenance: Copy-Specific Features of Incunabula (Leiden: Brill, 2025), pp. 51-80.

---

For more on the St Christopher woodcut and other early printed images in the John Rylands Library, see the recently released Early European Print collection on Manchester Digital Collections.

For more on John Bagford and the 18th century Apocalypse woodblock, see Ed’s previous blog post Experimenting with Print.

---