The Works of Geoffrey Chaucer (The Kelmscott Chaucer)

Finished eight weeks before William Morris's death, the Kelmscott Chaucer is the last and largest work of his private press: 556 pages of his own Chaucer type, framed in his own borders and initials, illustrated with 87 wood-engraved designs by his oldest friend, Edward Burne-Jones. It took roughly two years to print and, by most reckonings, four years to design, during which Morris's health was visibly failing; Burne-Jones wrote of hoping the book would be finished before Morris died. It very nearly was not.

The edition ran to 438 copies: 425 on handmade paper at 20 pounds each and 13 on vellum at 120 guineas, a price that a period source puts at roughly a year's wages for an ordinary worker on the cheap end and nine years' wages on the expensive end. Sotheby's, Christie's and Bonhams have all since sold copies for well into six figures in dollars.

This record concerns the copy held by the Iris and B. Gerald Cantor Center for Visual Arts at Stanford University, a gift of Mr. and Mrs. Nathaniel J. Kendall. Stanford also holds a separate Kelmscott Chaucer in its Green Library Special Collections, part of a complete run of all 53 Kelmscott Press titles funded in part by Herbert Fleishhacker; the two should not be confused, and this record treats only the Cantor Center's copy, illustrated here with its own Google Art Project photograph. Whether the Cantor's copy is paper or vellum is not stated in the institution's published catalogue description and could not be confirmed for this record.

What the book physically is

The Kelmscott Chaucer is a single folio volume of 556 pages, printed in double columns in black and red ink, its text set in the 12-point Chaucer type with headings in the larger Troy type, both designed by William Morris. It carries 87 wood-engraved illustrations designed by Edward Burne-Jones, 14 large ornamental borders, 18 smaller frames around the pictures, and 26 large decorated initial words, all designed by Morris and printed from wood blocks. Every opening was conceived by Morris as a single unified design rather than a page of text with pictures dropped in, which is part of why the book is often described as feeling more like an illuminated manuscript than a printed one.

The edition totalled 438 copies: 425 printed on handmade paper and 13 on vellum. Most paper copies left the press unbound or in a plain binding of quarter holland (linen) with blue paper-covered boards, supplied by the London binders J. and J. Leighton; a smaller run, most sources put the number at 48, was bound in full white pigskin over boards, tooled in blind to a design by Morris and made up at the Doves Bindery under T.J. Cobden-Sanderson, at an additional charge of 13 pounds 13 shillings per copy. The vellum copies were bound separately again. Stanford's Cantor Arts Center gives its own copy's overall dimensions, bound, as roughly 17 1/4 by 12 1/5 by just under 3 inches, but does not state in its published description which binding class the copy belongs to.

Two Oxford undergraduates and a fourteenth-century poet

The book's origin is not really in the 1890s at all. Morris and Burne-Jones met as undergraduates at Exeter College, Oxford, in the early 1850s, and their shared enthusiasm for medieval poetry and manuscripts, including visits to the Bodleian Library to look at illuminated books, dates from that friendship's earliest years. The Fitzwilliam Museum's own account of its copy of the Chaucer records that Morris referred to Chaucer as his master. Whatever else the Kelmscott Chaucer is, it is also the end point of a private enthusiasm the two men had shared for more than forty years before either of them tried to make a book out of it.

That the two men worked well together on this specific project is documented in their own words rather than inferred from the result. According to a widely reproduced letter, Burne-Jones worked on the Chaucer illustrations chiefly on Sundays, when Morris would come to his studio to talk and to read aloud while he drew. Of the finished book Burne-Jones wrote: if we live to finish it, it will be like a pocket cathedral, so full of design, and I think Morris the greatest master of ornament in the world. This is not the language of a man who felt his contribution was being swallowed by his collaborator's; if anything it credits Morris with more of the achievement than a modern eye, looking at 87 full-page Burne-Jones designs, might expect.

Designing an ideal book as an argument against industrial printing

Morris founded the Kelmscott Press in Hammersmith in January 1891 explicitly as a reaction against what mechanised, cheaply inked Victorian printing had done to the book as an object. In his own words, printed in his 1893 note on the ideal book, he began printing books with the hope of producing some which would have a definite claim to beauty, while at the same time they should be easy to read. He blamed the greyness of ordinary contemporary printing on cheap ink, and had a specially dense black ink shipped from the firm of Gebrueder Jaenecke in Hanover, Germany, to correct it. He commissioned the papermakers Joseph Batchelor and Son of Kent to make paper from linen rag on a mould, deliberately choosing a ribbed texture modelled on paper made in Bologna around 1473, and designed three watermarks of his own, a flower, an apple and a perch, for different sheet sizes. Vellum copies used a different supplier again, from the firms of Henry Band and William J. Turney and Company.

None of this was decorative choice divorced from argument. Morris's opposition to industrial capitalism, developed through the 1880s in his work for the Socialist League and in books like News from Nowhere, rested on the claim that mechanised production had made both work and its products ugly by stripping the maker's judgement out of the process. The Kelmscott Press, producing books by the same kind of hand labour, unified craft control and material choice that Morris associated with the medieval workshop, was conceived as a working demonstration of that argument rather than an exercise in nostalgia for its own sake. This is Morris's own stated position, argued at length in his lifetime; it is not a later reading imposed on the book.

From Burne-Jones's pencil to the printed page

Burne-Jones's contribution did not go from his hand to the woodblock in one step, and the intermediate steps are themselves a small, sourced story about credit. His illustrations began as delicate pencil drawings, and delicate pencil shading does not translate into a wood engraving, which reproduces only line. Emery Walker, the process engraver and a Kelmscott Press partner, photographed Burne-Jones's drawings as platinotype prints. The young artist Robert Catterson-Smith, hired by Morris for the purpose, then reworked each photograph, covering it with a wash of Chinese white and redrawing the image in bold, engraveable lines; his revised version, not Burne-Jones's original, was transferred photographically onto the woodblock, where William Harcourt Hooper cut it.

Catterson-Smith's name appears nowhere in the finished book. A specialist blog devoted to updating the modern census of surviving copies records the plain assessment that Morris was unwilling to acknowledge publicly his heavy reliance on photographic technology, and notes that the omission drew complaints even at the time. Of the 87 original Burne-Jones drawings, 85 are recorded as surviving today at the Fitzwilliam Museum in Cambridge, which also holds a number of Burne-Jones's preliminary sketches, so the difference between what Burne-Jones drew and what the book actually shows can, in principle, still be checked drawing by drawing.

Morris's failing health, and whether he knew

Accounts of exactly what was wrong with Morris in his last years do not agree with one another in the sources consulted for this record. One biographical summary dates the onset of serious kidney disease to 1891, the same year the Kelmscott Press was founded; other accounts describe him as diabetic, and still others cite tuberculosis or heart disease as contributing causes of his death. This record does not attempt to adjudicate the medical question; it notes that the secondary literature itself is unsettled on it, which is a finding worth recording rather than papering over.

What is better documented is that Morris kept working through the decline. In July 1896, with the Chaucer finished and the press still running down its remaining commitments, he travelled to Norway, a trip one account describes his doctor as understanding to be his last voyage; he returned with what is described as congestion of the lungs and continued dictating work into that August. Burne-Jones's own letters show he was thinking about his friend's mortality while the Chaucer was still in progress: in a passage recorded by one biographical source, he wrote of hoping the printing would not end before Chaucer, and a second planned Kelmscott book, an illustrated Malory's Morte d'Arthur, were both finished. The Morte d'Arthur was never published by the Kelmscott Press; none of its 53 issued titles carries that name, and whether the project was abandoned before or because of Morris's death is not detailed in the sources consulted here. Morris died at Hammersmith on 3 October 1896, not quite four months after the Chaucer's completion. One account attributes to Georgiana Burne-Jones, Edward's wife, the description of his death as coming as gently, as quietly as a babe who is satisfied drops from its mother's breast.

What the book cost, and what that meant

At publication in 1896 the 425 paper copies sold for 20 pounds each and the 13 vellum copies for 120 guineas, that is, 126 pounds; a copy specially bound in the Doves Bindery's pigskin cost a further 13 pounds 13 shillings on top of the paper-copy price. A specialist rare-book dealer's own published account converts the 20-pound price to roughly 100 contemporary US dollars and sets it against an average American wage of about 3.52 dollars a week, or 183 dollars a year, in the same period, putting the paper copy at a bit over half a year's average wages and the vellum copy at around nine years of them. That comparison is a dealer's own figure rather than one drawn from a peer-reviewed historical-wage index, and it compares an English retail price to an American wage; it is included here as a documented data point, not as a precise economic equivalence, and a reader wanting a rigorous period-wage conversion should treat it as a starting point rather than a verified figure.

By the standards of Morris's own enterprise the Chaucer was not a commercial success. One summary of the Kelmscott Press's finances describes Morris as drawing no salary from the press and personally covering shortfalls from the income of his textile and wallpaper businesses; smaller, cheaper Kelmscott titles are described as having subsidised the cost of producing large, elaborate volumes like the Chaucer. The book sold out, but it was expensive to make in a way that a subscription list alone did not fully cover.

After Morris died

The Kelmscott Press did not close the day Morris died. It continued publishing through 1897 and into 1898, completing an eight-volume edition of Morris's own Earthly Paradise and his epic poem The Story of Sigurd the Volsung, among other titles already in progress, and issuing A Note by William Morris on His Aims in Founding the Kelmscott Press as one of its last publications. Morris had asked Emery Walker and his secretary Sidney Cockerell to consider carrying the press on after his death; both men judged that the press was too much an expression of Morris's own personality to continue without him; it closed in 1898 having issued 53 titles in total.

Reputation followed almost immediately. Within a generation the Chaucer was being described as the most beautiful book of its age, and that judgement has held up remarkably well across more than a century of subsequent book design; auction cataloguers in the twenty-first century still routinely bill copies as, in one recent dealer's phrase, the finest book since Gutenberg. That kind of superlative is a marketing line as much as a scholarly one, but it reflects a consistent verdict rather than a manufactured one.

The market since: what copies actually make

None of the sales documented below is Stanford's own copy; no sale record connecting the Kendall gift to a prior auction or dealer transaction was found in the sources consulted, and the chain before the Kendalls owned it is undocumented. The figures here are comparanda: what the same edition, in different copies and different bindings, has made at public auction, given to show the shape of the market this copy sits inside.

In December 2024, Bonhams New York sold a paper copy as part of The Complete Kelmscott Press Collection of Joseph Mark Van Horn, a single-owner sale of one copy of each of the 53 Kelmscott Press titles. That Chaucer, rebound in near-contemporary full brown calf by Riviere and Son rather than in an original Kelmscott-period binding, and with recorded provenance from Goodspeed's Book Shop in Boston in 1963, made a hammer price of 89,800 US dollars against a high estimate of 90,000, for a total of 114,800 dollars with premium; the whole 53-volume collection totalled 690,000 dollars. In February 2024, RR Auction sold a different paper copy, in a later full white leather rebinding with a matching slipcase, for 90,750 dollars with premium included against an estimate of 65,000 dollars and up, without separately publishing the hammer figure. A further paper copy, in a binding closer to the original quarter-linen and blue-paper-boards style, made 68,750 dollars at a subsequent RR Auction sale; again the source consulted did not separate hammer from premium. At Sotheby's in London, a copy in one of the original 48 Doves Bindery pigskin bindings, with recorded provenance from a sale of 15 May 1941, carried a pre-sale estimate of 50,000 to 70,000 pounds; the result itself was behind a login wall and could not be retrieved for this record.

Two things stand out across this small sample. First, rebound copies, in leather bindings applied well after 1896, are common on the market and are not automatically worth less for it; several of the highest recorded prices above are for rebound copies. Second, an original 1896 pigskin Doves Bindery copy, of which roughly 48 were made, appears to carry a meaningfully higher estimate than a plain or later-rebound paper copy, though the one such copy identified here did not have a retrievable hammer result to confirm that with a completed sale.

Stanford's two Kelmscott Chaucers, and which one this is

Stanford University holds Kelmscott Press material in at least two separate places, and the two should not be treated as the same object. Stanford's Green Library Special Collections built a complete run of all 53 Kelmscott Press titles beginning with an initial gift of six titles in fifteen volumes in 1926 to 1927; the purchase of a copy of the complete Chaucer was authorised as part of that gift by the donor Herbert Fleishhacker, but by the account consulted here took nearly fifteen years to actually fulfil, meaning that Special Collections copy was likely not acquired until sometime around the early 1940s.

The copy this record describes is a different one: it is held by the Iris and B. Gerald Cantor Center for Visual Arts, Stanford's art museum, and its own published object record credits it as a gift of Mr. and Mrs. Nathaniel J. Kendall, with no connection stated to the Fleishhacker-funded Special Collections copy. This is the copy reproduced in the Google Art Project photograph used as this record's lead image, and the one credited to the Cantor Center on that photograph's Wikimedia Commons page and file category. Beyond the donor's name and the museum's own overall dimensions for the bound volume, no further detail, no accession number, no binding description, no statement of paper or vellum, and no acquisition date, was found published for this specific copy in the sources consulted. The Cantor Arts Center's own searchable collections database exists online but requires passing a CAPTCHA verification step to return results, and could not be queried directly for this record.