Manuscript Road Trip: You Can’t Argue with Science!

The Flight into Egypt, Walters Art Museum, MS W.188, f.112r

This is a story about how science and forensics can help translate a question of subjective connoisseurship into one of objective certainty. In other words, you can’t argue with science.

Back in 2001, I was hired by the Department of Special Collections at Wellesley College (in Wellesley, Mass.) to catalogue their very fine collection of pre-1600 manuscripts. I spent several years working my way through the collection, researching contents, origin, provenance, and physical characteristics of the codices, cuttings, and single leaves, eventually producing a 91-page document that became the source for the library’s MARC records. Some of manuscripts have been digitized and are online here. Five were included in the Beyond Words exhibition in 2016; those records are here.

Wellesley College, MS 29, f. 14v (St. Christopher)

Today, I want to focus on one manuscript in particular, Wellesley MS 29. MS 29 is a late-fifteenth/early-sixteenth-century Book of Hours for the Use of Utrecht, written and decorated in the Northern Netherlands. The full-page miniatures in the manuscript have recently been attributed by James Marrow to the “Masters of the Dark Eyes” (see Hamburger, et al., Beyond Words: Illuminated Manuscripts in Boston Collections (Chestnut Hill: McMullen Museum of Art, 2016), Cat. 128 (p. 163) – you can download Marrow’s entry and description of the manuscript from the Beyond Words website). Artistic attribution is a tricky and often somewhat subjective business. This is particularly true when dealing with medieval manuscripts, whose artists are rarely named and may instead be identified as the Master of This or the School of That. That kind of connoisseurship is beyond the scope of my training and expertise – I can tell that MS 29 is Netherlandish from the general style, but I leave such specific attributions to expert art historians like James Marrow.

Another such expert with an impressive visual memory and a long track record of artistic attribution is Peter Kidd, a British manuscript consultant whom I have mentioned before. Several months ago, he wrote to Ruth Rogers, the Curator of Special Collections at Wellesley, with some questions about MS 29. In particular, he thought that while the full-page miniatures seemed appropriately dated and attributed, something seemed “wrong” about the small historiated initials, a selection of which are seen below.

 

MS 29, f. 21r

MS 29, f. 40r

MS 29, f. 79r

A trio of suspicious miniatures…

Kidd wrote about his suspicions on his blog, Medieval Manuscripts Provenance, in May 2019 (he also had concerns about Wellesley MS 27, but I’m going to focus on MS 29 today). By “wrong,” he meant that he suspected the historiated initials may have been later additions, nineteenth-century forgeries added to increase the value of the manuscript. Kidd argued that the initials may even have been produced by the Spanish Forger (about whom more here). They seemed all right to me when I catalogued the manuscript in 2001, but as I’ve said, I’m no art historian. Previous studies hadn’t expressed suspicion about the initials, but Kidd thought they might be forged. It’s a classic art historian standoff, one opinion vs. another. Fortunately, science can help turn opinion into fact. In this case, the question was settled by the application of X-Ray Fluorescence (XRF).

X-Ray Fluorescence uses an x-ray beam to produce a chemical signature, identifying the elements that make up the sample being tested. In the case of a medieval manuscript, XRF testing can determine the chemical composition of mineral-based pigments. Because some chemical signatures can be identified as artificial man-made pigments, as opposed to the simple mineral pigments used in the Middle Ages, XRF testing can help determine whether pigments are authentic or are later additions. For example, the “Paris Green” that gave the Spanish Forger away, also known as “Scheele’s Green,” is a man-made pigment first synthesized in 1775 whose chemical signature shows a distinctive arsenic spike in addition to the expected, and perfectly normal, copper signature.

After hearing from Peter Kidd, Ruth Rogers decided to investigate further. She invited me to join her and Wellesley book conservator Emily Bell on a field trip to the Boston Museum of Fine Art’s digital lab to observe as they conducted XRF testing on MS 29. Testing was conducted by Michele Derrick and Richard Newman using a Bruker Artax open architecture spectrometer for 120 seconds at 40 kV and 700 microamps. I am very grateful to both of these skilled technicians for their careful testing and thorough reporting. It was a fascinating and truly educational afternoon.

 

This slideshow requires JavaScript.

Together, Ruth and I selected pages and specific targets, focusing on reds, blues, and greens. Miniatures that were likely original were scanned as a baseline and those that were suspect were scanned for comparison. The results could not have been more clear.

The green pigment in the full-page miniatures that were tested (St. Christopher (see above) and The Man of Sorrows, f. 78v) was defined by a spike in copper only, no arsenic. So nothing suspicious there.

Chemical signature of the green pigment on the king’s robe in the upper right corner of f. 78v: the combination of chemical elements (with a spike in [Cu], copper) suggests that the pigment was ground from malachite, perfectly normal for a medieval green.

But the historiated initials that felt “wrong” to Peter Kidd? His “wrong” instincts turned out to be absolutely right. The green in the three historiated initials that were tested (ff. 21, 40, and 79) showed a clear and dramatic spike in the arsenic [As] signature in addition to copper. The pigment was copper-acetoarsenite, also known as Paris Green. As Kidd suspected, they were all nineteenth-century forgeries. You can’t argue with science.

Chemical signatures for the green pigment in three of the suspect historiated initials (top to bottom, ff. 79r, 21r, and 40r). The arsenic spike is outlined in black.

The evidence suggests that the spaces for these initials were originally left blank, that is, they were planned but never painted. In other words, there was no evidence of scraping or overpainting. Our forger, whoever s/he was, added the historiated initials to the blank spaces in the nineteenth century, probably to add value to the manuscript. There are certainly many other manuscripts out there that have been supplemented in this way but haven’t yet been studied carefully enough to arouse suspicion. Unfortunately, not everyone has access to an XRF lab, and so such suspicions usually must go untested. But there are professional labs that can do this work along with other kinds of material testing, and the results are almost always worthwhile, adding scientific evidence to artistic judgement. Even the Voynich Manuscript (you knew I’d bring that up, didn’t you?) has been subjected to material testing, to determine whether it might be a modern forgery. The results of testing on Voynich ink and pigments showed nothing to indicate that the manuscript ISN’T medieval, and carbon-14 testing dated the manuscript’s parchment to the early fifteenth century.

 

“Paris Green” can’t hide from XRF

Given the clear results of the XRF testing, Wellesley will need to update the manuscript’s digital records to indicate that some of the initials are nineteenth-century forgeries, adding a fascinating additional chapter to the history of this already-intriguing manuscript. The story of Wellesley MS 29 shows how manuscript studies can benefit from the combination of modern forensics and knowledgeable connoisseurship. There is always room for new understanding of old books.

Also, trust your gut.

And never argue with science.

Manuscript Road Trip: Australian Edition

The Flight into Egypt, Walters Art Museum, MS W.188, f.112r

I know it’s been a month since I last posted, but I have a good excuse. I’ve been in Australia for several weeks, visiting my daughter (who is studying abroad there this term) and exploring antipodal medieval manuscripts online and in person. I was invited to deliver a lecture at The Medieval and Early Modern Centre at the University of Sydney, and this post is based on that lecture.

Collections of medieval and Renaissance manuscripts in Australia and New Zealand were primarily developed by bequest, given by British collectors. As such, most of the manuscripts have a direct British provenance and didn’t come through the hands of dealers we have come to know in our explorations of North American collections (such as Duschnes, Ege, Ferrini, Rosenthal and Kraus, among others). Even so, while exploring online collections in Australia and New Zealand, I was able to identify later acquisitions of single leaves that have connections to single leaves in North America. Welcome to the Manuscript Road Trip: Australian Edition.

Sydney Harbor at sunset (photo by me)

Australia and New Zealand are fortunate in that detailed surveys have been undertaken of the national corpus of medieval manuscripts by such esteemed experts as Christopher de Hamel, K. V. Sinclair, Margaret Manion, Vera Vines, and Alexandra Barratt, among others. There are dozens of pre-1600 European manuscripts in antipodal collections, many of which have been imaged and made available online.

Spanish Forger cutting, University of Sydney, Special Collections, s.n. (Voelkle Catalogue L134)

Digitization projects are underway at the University of Sydney, where images have been posted online from several manuscripts and a handlist can be found here. A Parisian Book of Hours (RB Add. Ms.58) is noteworthy, as is the University’s Spanish Forger cutting (at left). The spectacular manuscript collection at the State Library of Victoria has been digitized as well, and many Australian collections have added records, if not images, to the national Trove database (try searching for keyword “manuscript,” with date limits of 1000 – 1600, limited to Australian collections).

As my readers will have noticed, I am always on the lookout for single leaves that can be traced back to Otto Ege. Lo and behold, there is indeed a previously unrecorded “Original Leaves from Famous Books: Eight Centuries” set at the State Library of Queensland in Brisbane. This particular portfolio is comprised of printed and manuscript leaves: the handwritten pages come from a thirteenth-century Bible, a fourteenth-century Aristotle, and a fifteenth-century copy of Livy’s History of Rome. These are not the most visually impressive group. Ege didn’t select them any art historical “wow” factor. He used them here as examples of the development of lettering, as calligraphic specimens of Gothic bookhand, Gothic cursive, and Humanistic bookhand:

Original Leaves from Famous Books: Eight Centuries, Leaf 1 (Bible) (State Library of Brisbane, RB 686.23 19– )

Original Leaves from Famous Books: Eight Centuries, Leaf 2 (Aristotle) (State Library of Brisbane, RB 686.23 19– )

Original Leaves from Famous Books: Eight Centuries, Leaf 3 (Livy) (State Library of Brisbane, RB 686.23 19– )

Based on my survey of Sinclair’s catalogue of Australian collections and additional information kindly sent to me by U. Sydney Professor Nick Sparks, I count well over 100 single leaves in Australia. A few have been imaged and are online, for example, these University of Tasmania leaves from two different Books of Hours. At the University of Sydney, there are several dozen single leaves, all of which are worthy of study and some of which may be connected to known leaves elsewhere. In 1967, Sinclair identified four leaves in private Sydney collections as having come from the same fifteenth-century French Breviary: a bifolium in collection of Sir Edward Ford (Sinclair no. 67), and one each in the collections of J. M Ward (Sinclair no. 83) and R. I. Jack (Sinclair no. 85). Sinclair’s descriptions of single leaves are extremely useful because they provide not only the date and place of origin but also the dimensions and the number of lines per page, objective data that can help identify leaves that came from the same parent manuscript. The data for the four breviary leaves is consistent with a manuscript dismantled by Otto Ege in the 1940s and now used as leaf 24 in his “Fifty Original Leaves” set. I haven’t confirmed this identification yet, but I am ever hopeful. Once collections begin digitizing their leaves, more such connections will surely be found. I am certain there are more leaves of this gorgeous fourteenth-century English antiphonal out there somewhere:

 

Antiphonal (leaf), England, early 14th century (Geelong Church of England Grammar School, Corio, Victoria) (Manion and Vines no. 42, pl. 25)

Now let’s head down to New Zealand and see what we can find there.

Auckland Libraries, Sir George Grey Special Collections, 1480 BIBL

Munich, CLM 29260

As in Australia, the medieval manuscripts in New Zealand have been expertly catalogued, by native New Zealander Christopher de Hamel and by Alexandra Barratt, among others. A forthcoming study by Alexandra Barratt of leaves in Auckland will focus on important fragments still being used as binding scrap. At left is one such example, hiding in the gutter of an early printed book, an incunable. These narrow strips of recycled parchment are called binding stays, and they serve to protect paper from tearing as the gatherings are sewn together. These hidden treasures are difficult to photograph, and difficult to work with, but the payoff for that effort can be enormous. Barratt has confirmed that these ninth-century binding stays come from the same manuscript as the fragment above, from the Bayerische Staatsbibliothek in Munich. You can read more about the project at http://medievalbookbindings.com/.

Dunedin Public Library, Reed Ms Frag. 13

In the 1989 catalogue, Medieval and Renaissance Manuscripts in New Zealand Collections, de Hamel draws several connections between leaves in New Zealand and known leaves elsewhere. One such is Dunedin Public Library Reed Fragment 13 (right), a leaf of a lovely English Bible sometimes known as the Bohun Bible, of which several dozen other leaves are known in collections in Europe, England and the U.S. including Houghton MS Typ 966 at Harvard University, right around the corner from my office.

Dunedin Public Library, Reed MS Frag. 17

Also of note are the stunners collectively known as Dunedin Frag. 17, three leaves from a well-known Book of Hours attributed to the Boucicaut Master. Here’s a leaf from the same manuscript that I illustrated in my blog just a few weeks ago, from the University of North Carolina.

De Hamel also identifies two leaves in the Dunedin Public Library as having come from manuscripts known to have been dismembered by Otto Ege (more on this below). While looking through the New Zealand catalogue, however, the description of Dunedin Frag. 42 caught my eye: it is a lectionary from twelfth-century Italy of 22 lines per page that sounds a lot like the Ege lectionary we have met before. The dimensions are spot on as well. But without images, there’s no way to be absolutely sure.

Fortunately, the Dunedin library has recently imaged all of their codices and leaves and put them on a Flickr site accompanied by de Hamel’s metadata. Flickr is an easy and popular platform for presenting digitized manuscripts, in part because it allows for comments and crowdsourced cataloguing, a strategy that has been very successful for small collections in the U.S that don’t have the storage space or software to post the images on their own servers. Calling up the online image of Reed Fragment 42 confirms that it is indeed a leaf of Ege’s lectionary, “Fifty Original Leaves” (FOL) no. 3.

Dunedin Public Library Reed MS Frag. 42

Now that the Dunedin fragments have been digitized and posted online, I can also confirm de Hamel’s identification of Reed Fragment 7 and Reed Fragment 51 as having come from the same manuscript as Ege leaves FOL 9 and FOL 2. The core collection was donated to the Dunedin Public Library in 1948 by Alfred Hamish Reed (1875 – 1975) along with a sizable bequest that has allowed the library to make more recent purchases, including Frag. 42 (purchased from Maggs Brothers in 1982) and Frag. 51 (purchased from Quaritch in 1985). Fragment 7 was purchased during Reed’s lifetime from W. A. Stewart in Morrinsville, Pennsylvania, in June 1952.

Also worth noting is that the Dunedin Reed Frag. 33 (below left) is from the same manuscript as one of the leaves at the University of Tasmania (below right). That’s quite a co-incidence. The original manuscript – a mid-fifteenth-century Book of Hours – belonged to a San Francisco collector named John Francis Neylan, who died in 1960. It was sold at Sotheby’s on 18 June 1962 (lot 121) to a London outfit called The Folio Society that was famous for dismembering books to sell as individual leaves or leaf-collections. The Dunedin library purchased their leaf from The Folio Society in 1963; it is unclear when or how the University of Tasmania acquired theirs. And in fact, there may be yet another leaf from this manuscript in Australia, at least there used to be one, at the University of New South Wales; Sinclair’s catalogue describes their MS 2 as a Book of Hours with identical dimensions to these and the same number of lines (Sinclair no. 18). Unfortunately, like many of the University of New South Wales manuscripts described by Sinclair, this leaf is now lost, so I cannot as yet confirm this identification.

Dunedin Public Library, Reed Frag. 33

Univ. of Tasmania, s.n.

I leave you with an image that has nothing to do with manuscripts but everything to do with Australia: “Self-portrait with wallaby.”

See you in West Virginia!

Manuscript Road Trip: The Spanish Forger

The Flight into Egypt, Walters Art Museum, MS W.188, f.112r

One of the most rewarding things about this virtual road trip I’m on is learning about collections that are new to me. A comment on last week’s post pointed me towards the manuscripts owned by Reed College in Portland, Oregon, among which is a leaf illuminated by the Spanish Forger. This reminded me that I promised several months ago to devote a post to the Forger, and before I delve into the riches of California’s manuscript collections, I thought now would be a good time to get off the proverbial highway and follow through.

I first encountered the Spanish Forger in 1997 while working for private collector Lawrence J. Schoenberg, whose collection is now part of the Schoenberg Institute for Manuscript Studies at the University of Pennsylvania. At that time, he owned the piece shown below, a framed miniature of a group of young nobles playing chess.

Univ. of Pennsylvania, LJS MS 33

I was immediately enchanted and amazed, both by the delightfully saccharine setting and by the Forger’s sheer nerve in attempting, and for some time pulling off, such a massive deception. I’ve been a fan ever since.

The Spanish Forger painted numerous miniatures and panels in a late-medieval style at the end of the 19^th and into the 20^th century. No one knows who he was, but given the number of leaves and panels attributed to him, he must have made a significant amount of money selling his forgeries to an unsuspecting audience.

The Betrothal of St. Ursula (The Morgan Library and Museum, New York)

The story of the Spanish Forger begins with this panel painting of “The Betrothal of St. Ursula,” a painting that had been ascribed to fifteenth-century Spain on stylistic grounds. In 1930, Belle da Costa Greene (a prominent art historian who had been J. P. Morgan’s private librarian and who was one of the first women to be elected a Fellow of the Medieval Academy of America) refused to support its purchase for New York’s Metropolitan Museum because she suspected it was a forgery. Greene’s life story, by the way, is pretty extraordinary. Her father, Richard Theodore Greener, was the first African-American to graduate from Harvard, and Belle lived her life as a black woman “passing” and moving easily among the New York elite in the early twentieth century.

It was Greene – at that time curator of the Pierpont Morgan Library – who first identified the Spanish Forger’s distinctive characteristics and gave him his name. After the St. Ursula panel  was later tested using neutron activation analysis, it was discovered that the green pigment in the painting was copper arsenite a.k.a. Paris Green, which was not available before 1814, confirming Greene’s suspicions.  Because French newsprint has been found behind some of his panels, it is suspected that he actually worked in Paris, but the name Greene gave him has stuck. In 1988, the painting that gave the Spanish Forger his name was given to the Morgan Library and Museum.

To his credit, the Forger always worked on medieval parchment or panels. He scraped paint from wood or text from parchment, painting forged miniatures on authentic supports, knowing full well that – in the case of manuscripts – illustrated leaves were much more valuable than unillustrated. While preparing a major exhibit of the Forger’s work for the Morgan Library in 1978, curator William Voelkle discovered that the Forger had based much of his work on illustrations found in five volumes on medieval and Renaissance life and culture by Paul Lacroix (published between 1869-1882); many of his paintings are clear copies of Lacroix illustrations, as shown here.

The Forger’s style is easy to recognize. His faces often have a pronounced tilt, certainly, but one feature is particularly distinctive; the ladies in his noble settings display a prominent décolletage that would never have been found in an authentic medieval painting. Technical features of his work also give him away.

The Morgan Library and Museum, MS M. 786c

In a genuine medieval manuscript, the gold leaf would have been applied before the colors. The Forger tended to apply his gold as a final step; a close examination of his work often finds gold overlapping the color rather than the (correct) other way around.

NY, Columbia Univ., Plimpton Add. MS 18

All of these features – head tilt, cleavage, and gold-overlap – are evident in the miniature at left, a cutting in the collection at Columbia University. At first glance, this miniature appears to be an authentic medieval scene of three nobles playing chess in a courtly setting. But if you turn the piece over, the Forger’s cleverness in using medieval parchment betrays him.

NY, Columbia Univ., Plimpton Add. MS 18 (dorse)

The dorse (shown below) is a snippet of liturgy from the celebration of Pentecost, a text that most certainly would never have been illustrated by a courtly scene of lords and ladies playing chess. This disconnect is typical of the Forger’s work, as the subjects of his miniatures rarely relate to the text they accompany. In addition, the vertical text in the lower right corner is a catchword, indicating that, if the leaf were whole, the miniature would have been found in the extreme lower left-hand corner of the other side, a placement totally at odds with medieval page-layout practices.

Voelkle’s catalogue of the 1978 exhibit includes a handlist of more than one hundred works attributed to the Forger (see W. Voelkle, The Spanish Forger). He has recently updated this catalogue raisonné to include well over two hundred identified works (see “The Spanish Forger: Master of Manuscript Chicanery” in Thomas Coomans & Jan De Maeyer, eds., The Revival of Medieval Illumination: Nineteenth-Century Belgium Manuscripts and Illuminations from a European Perspective (Leuven, 2007, pp. 207-227)). The Forger essentially flooded the market with leaves and miniatures at a time when collectors were clamoring for such things. He must have made a fortune.

Spanish Forger Road Trip

In addition to Columbia University, the Morgan Library and Reed College, the Forger’s work can be found in many other public U. S. collections. In fact, you could make your way from East Coast to West on a Spanish Forger road trip alone, visiting New Hampshire (Dartmouth College),  Massachusetts (Harvard University, the Boston Athenaeum, Wellesley College and the Museum of Fine Arts, Boston), Connecticut ( Yale University), New York (Columbia University, the Metropolitan Museum of Art and the Morgan Library and Museum), Pennsylvania (the University of Pennsylvania and the Free Library of Philadelphia), Maryland ( the Walters Art Museum), Ohio (the Allen Memorial Art Museum at Oberlin College and the Cincinnati Art Museum), Michigan (Michigan State University), Texas (the Harry Ransom Center at the University of Texas at Austin), Colorado (the University of Colorado, Boulder), and Oregon (Reed College). I’m sure I’ve missed a few. Ironically, the Forger’s works are now collectibles in their own right and sell for thousands of dollars; recent sales at Christie’s and Sotheby’s attest to the Forger’s popularity, and there are many U. S. collectors and dealers who own examples of his work. I’d love to own one of his paintings myself.

Last week, we explored manuscripts in Washington and Oregon. Let’s get back on I-5 now and make a quick stop at Reed College in Portland before we turn south to California. The cuttings at Harvard and Columbia as well as the leaf recently sold at Sotheby’s are all closely related to the Spanish Forger miniature owned by Reed  (shown below). The Forger made use of several dozen leaves from this large early-fifteenth-century Italian antiphonal, retaining the original borders but replacing the top few lines of music on each leaf with a secular scene (see Voelkle p. 75, “choirbook stock A”).  The Reed leaf was L18 in the Voelkle catalogue (fig. 234), at which time it belonged to Harry A. Levinson of Beverly Hills, California. Reed acquired the leaf from Bloomsbury Auction House in 2008.

Reed College Special Collections, BX875.A2 S63

The Reed College leaf preserves the antiphon “Quidcumque ligaveris,” a piece sung during the celebration of feasts of St. Peter. The illustration is completely at odds with the text, showing instead the triumphant entry of Crusaders into Jerusalem. The Spanish Forger was marketing to an early twentieth-century audience that was more interested in Gothic illustration than Latin text. His commercial success depended on buyers who wouldn’t notice the text/image disconnect or the stylistic inconsistencies.

He didn’t count on Belle da Costa Greene.