"Tahmuras Defeats the Divs", Folio 23v from the Shahnama (Book of Kings) of Shah Tahmasp

A single leaf, folio 23 verso, from the grandest illustrated manuscript of Persia's national epic ever made: the Shahnama produced for Shah Tahmasp I in the royal Tabriz workshop between about 1522 and 1535. The painting, attributed to Sultan Muhammad, shows the legendary king Tahmuras defeating the divs and winning the art of writing from them as ransom, one of a cluster of early folios in which this painter staged the poem's founding-kings sequence as an argument for civilization arriving through royal conquest, at the very moment a new dynasty's workshop was asserting its own authority.

The manuscript was sent as a diplomatic gift to the Ottoman Sultan Selim II in 1568, resurfaced in the Rothschild collection in Paris by 1903 after an undocumented gap of centuries, and was sold to the American collector Arthur A. Houghton Jr. in 1959. Houghton gave this folio to the Metropolitan Museum in 1970 as part of a larger gift of 78 illustrations, then spent the following two decades selling the remaining folios individually, an act described by one journalist as an extraordinary act of vandalism, before his heirs exchanged what was left of the physical manuscript with the Iranian government in 1994 for Willem de Kooning's Woman III. Folios from the same book are now scattered across at least a dozen public collections and the private market, where a single leaf has sold for more than seven million pounds.

This folio's own story since 1970 is the simplest part of that history: it has never left the Metropolitan Museum, the one piece of a dismembered masterpiece that was never put back on the market at all.

What the folio is

This is a single leaf from a manuscript, not a picture that was ever meant to be seen alone. The painting occupies roughly two-thirds of the sheet, 28.3 by 18.6 centimeters, set into a full folio of 47 by 32.1 centimeters whose remaining surface carries Ferdowsi's Persian verse in nasta'liq script, framing the image on the recto and continuing on the verso. The support is paper, worked in opaque watercolor with ink outlining and passages of silver and gold, the silver now oxidized in places to a dull grey where it once read as bright metal, a change common across manuscripts of this class and period.

The composition shows a mounted king galloping across a meadow strewn with rocky outcrops, spotted and mottled demons cowering, fleeing, or already bound around him. It is one opening of a manuscript that originally ran to 759 pages with 258 such illustrations, of which this is folio 23 verso, meaning it falls very early in the physical sequence: the sixth of the manuscript's opening cycle of legendary kings.

The scene: Tahmuras and the gift of writing

The subject is Tahmuras, the third king of Ferdowsi's mythic Pishdadian dynasty, known in the poem by the epithet div-band, the demon-binder. Ferdowsi's narrative has Tahmuras make war on the divs, the demons of Iranian myth, and defeat them in battle. Rather than kill his captives outright, he accepts a ransom: in exchange for their lives, the divs teach him the art of writing, an origin story the poem extends to thirty scripts, among them Arabic, Chinese, Greek, Persian, and Sogdian, so that Tahmuras becomes, in the epic's telling, the point at which human beings first receive the alphabet.

Sultan Muhammad's picture catches the king mid-gallop rather than at a moment of static triumph, his horse crossing broken ground scattered with cowering figures whose faces the painter has pushed toward caricature: mottled, spotted skin, gaping mouths, and expressions closer to comic terror than to menace. The Metropolitan Museum's own published description of the sheet singles this out specifically, noting that the humor of the divs' ghastly faces and gestures, and the painterly treatment of their spotty skin, are typical of the artist's style. It is a scene about civilization acquired by force from a defeated supernatural order, and the painter has chosen to make its villains ridiculous rather than terrifying, which is itself a reading of the text.

Where it falls in the manuscript: a dynasty's opening argument

The Shahnama's narrative opens with a sequence of legendary kings who, in Ferdowsi's telling, are less historical rulers than personifications of the stages by which human civilization was built: Gayumars, the first man and first king; Hushang, who discovers fire by accident and founds the fire-festival of Sada in celebration; Tahmuras, who wins writing from the divs; and Jamshid, who is credited with the crafts, the calendar, and the division of society into orders, before the cycle turns dark with the tyrant Zahhak.

In this particular manuscript, that sequence receives outsized pictorial weight at the very point a new royal workshop was establishing its own authority. The folio immediately before this one, folio 22v, shows Hushang's Feast of Sada, the fire-discovery scene; folio 23v shows Tahmuras winning writing; both are attributed to the same painter, Sultan Muhammad, and both entered the Metropolitan Museum in the same 1970 gift as consecutive numbers, 1970.301.2 and 1970.301.3. A few folios later, folio 29v shows Zahhak being told the prophecy of his own downfall, also Sultan Muhammad's hand. Read in sequence, the manuscript's opening does not merely illustrate Ferdowsi's poem; it stages, for a newly established Safavid dynasty commissioning the grandest copy of the national epic ever made, an argument that kingship itself is the vehicle of civilization, arriving stage by stage through combat, sacrifice, and the taming of chaos. This record follows the discipline of grading such readings: the parallel between Tahmuras winning writing from the divs and a young dynasty asserting its own civilizing legitimacy is an argued interpretation available from the sequence and subject matter as documented, not a statement of what Sultan Muhammad or his patrons wrote down as intent, which is not recorded.

The royal workshop that made it

The manuscript was produced in the royal kitabkhana, the combined library and design atelier of the Safavid court, in Tabriz, beginning under Shah Ismail I and continuing under his son and successor Shah Tahmasp I, who inherited the project as a young prince and appears to have taken a close personal interest in it as an adult ruler. Work is generally dated across the years 1522 to 1535, though production and completion of the whole manuscript, including its text and binding, is sometimes extended in the literature to as late as 1537 or beyond.

The kitabkhana at its fullest brought together an estimated fifteen painters alongside calligraphers, illuminators, book-binders, and the specialist gilders, margin-rulers, and burnishers whose labor is rarely named at all in the historical record. Named painters associated with the project include Sultan Muhammad, Mir Musavvir, Aqa Mirak, Mir Sayyid Ali, Dust Muhammad, and, joining later, Mirza Ali, Sultan Muhammad's own son. Sultan Muhammad is generally credited with directing the workshop for approximately its first hundred folios, the section into which this folio falls, before oversight of later sections passed to Mir Musavvir and Aqa Mirak. No contract, payroll, or workshop ledger for the manuscript survives in the sources consulted; the number and division of hands is a scholarly reconstruction built from style, not from a document that names who was paid to do what.

Sultan Muhammad's hand

Sultan Muhammad is the single artist most closely associated with this manuscript's most celebrated pages, including the Court of Gayumars, folio 20v, now in the Aga Khan Museum in Toronto and widely described in the literature as the outstanding picture of the whole book. That folio is a different object from the one described in this record, made by the same hand for the same manuscript, and it must not be read as a picture of this folio's subject; the two should never be conflated.

What connects them is a single painter's opening run through the manuscript's founding-kings sequence. The attribution of this folio to Sultan Muhammad is not disputed in the sources consulted, and it sits inside a small, internally consistent group: the Feast of Sada immediately preceding it, this folio, and Zahhak is Told His Fate six folios later, all catalogued by the Metropolitan Museum to the same hand, all part of the same 1970 gift, and all occupying the same early narrative position that later scholarship assigns to Sultan Muhammad's direction of the workshop. The attribution is therefore not resting on this folio in isolation but on a cluster of stylistically and physically related sheets that entered the same collection together.

The scholarship: Welch, Dickson, and what came after

The foundational modern study of this manuscript is a direct product of its own dispersal. Stuart Cary Welch, then a curator at the Fogg Art Museum and later at the Metropolitan, published A King's Book of Kings in 1972, timed to the 2,500th anniversary of the Persian monarchy and drawing heavily on the folios Houghton had just given the Met, with a checklist of all 258 known paintings. Welch went on to co-author, with the Princeton historian Martin Bernard Dickson, the fuller scholarly reckoning: The Houghton Shahnameh, two volumes published by Harvard University Press in 1981 in a limited run of 750 copies, running to well over 800 pages combined with color and monochrome collotype plates. It remains the standard reference for the manuscript's chronology, the identification of individual painters, and the internal evidence for how the workshop was organized, and it is the source this record leans on most heavily for the workshop's structure and Sultan Muhammad's role within it.

Sheila Canby, later the Met's own curator of Islamic art, has written on the manuscript repeatedly since, including a reduced-format study for the Museum published through Yale University Press, generally affirming the Dickson and Welch narrative while adding archival and historical context drawn from Ottoman and Safavid sources unavailable or underused in 1981. More recent scholarship, including a 2016 paper by Abolala Soudavar explicitly framed as a defense of Dickson and Welch against several papers attempting to revise their conclusions, indicates the manuscript's attribution history is still an active scholarly argument rather than a settled catalogue, though the disputes concern chronology and hands elsewhere in the book rather than a challenge to the attribution of this specific folio.

A gift to the Ottoman Sultan, 1568

The manuscript left Safavid royal hands within a generation of its completion. It was sent to Istanbul as a diplomatic gift to the Ottoman Sultan Selim II in 1568, part of a much larger consignment: Ottoman records describe its arrival at the Iranian embassy in February 1568 accompanied by thirty-four camels of other gifts, a scale that marks the manuscript as one item within a formal state exchange rather than an isolated present. It then remained within the Ottoman royal library at the Topkapi Palace for an extended period; a chain reported through Metropolitan Museum catalogue material carries it further, from Selim II's custody to that of Sultan Selim III by around 1800, though this record was not able to independently verify that second Ottoman step against a directly consulted Museum page, and it is flagged accordingly.

Why a masterpiece of a rival dynasty's royal workshop was sent as a diplomatic gift, and what political message the gift itself carried between two states that had fought repeated wars across the sixteenth century, is a question this record does not attempt to resolve; the sources consulted record the fact and date of the gift without analysing Ottoman-Safavid diplomatic intent behind it.

An undocumented passage to Europe

Between an Ottoman royal library and its first securely documented appearance in a private European collection lies a gap the sources consulted do not close. The manuscript surfaces in the collection of Baron Edmond James de Rothschild in Paris by 1903, a full three and a half centuries after the 1568 gift and against the backdrop of the Ottoman Empire's political decline in the same period. No source consulted for this record documents how, when, or through whom the manuscript passed out of Ottoman custody and into the international art trade, or how it reached the Rothschild family specifically. That is a genuine, extended gap in the chain, not a smoothed-over transition, and it is recorded as one below rather than bridged with a guess.

The Rothschilds, and the sale to Houghton

Once in Rothschild hands the descent is more legible. Baron Edmond James de Rothschild held the manuscript in Paris from by 1903 until his death in 1934, when it passed to his son Baron Maurice de Rothschild, who kept it in Paris and at Pregny near Geneva until his own death in 1957; a further brief passage to Baron Edmond Adolphe de Rothschild is reported in the catalogue chain consulted, covering 1957 to 1959. On the dispersal of the Rothschild manuscript holdings that followed Maurice de Rothschild's death, the book was sold in New York in November 1959 through the dealers Rosenberg and Stiebel to the American collector and bibliophile Arthur A. Houghton Jr., a Steuben Glass executive, trustee and later president of the Metropolitan Museum, and a major manuscript collector in his own right as founder of Harvard's Houghton Library. A reported purchase price of 350,000 dollars appears in one source consulted, The Critic magazine's 2024 account of the manuscript's history; this record treats that figure as reported rather than independently confirmed against a primary sale record, and flags it for a second source.

The decision to take the manuscript apart

Houghton owned the complete manuscript for eleven years before making the first move that would end its physical unity: in 1970 he gave 76 folios, carrying 78 of the manuscript's 258 illustrations, to the Metropolitan Museum, this folio among them, formalized in the credit line the Museum still carries today, Gift of Arthur A. Houghton Jr., 1970. Structured correctly, a gift of this scale to a public institution functions as a form of tax relief for the donor as well as a benefaction to the Museum; the sources consulted describe the later individual sales explicitly in tax terms, and there is no reason given in any source consulted to read the 1970 gift differently.

Houghton kept the remainder, by most accounts around 118 to 120 illustrations plus the text and binding. In 1974 he attempted to sell what remained as a single intact unit, with an asking price reported at 28.5 million dollars; no buyer emerged. Following that failed sale, Houghton began disposing of individual folios one at a time, by outright sale and by further gift, over the following two decades. Christie's in London auctioned seven folios on 17 November 1976 for a combined 785,000 pounds, with the single best folio making 280,000 pounds, a price reported at the time as a record for any work of Persian or Islamic art. A second Christie's sale in October 1988 dispersed fourteen further folios for a combined 986,800 pounds, individual lots ranging from 14,300 to 253,000 pounds. The journalist Michael Prodger, writing in The Critic magazine in 2024, characterizes the whole course of action starkly, describing Houghton's 1974 decision as an extraordinary act of vandalism; a separate account, published on the manuscript-history blog The Books In My Life, states its author's position just as plainly: I remain disgusted by the destruction of a unique, precious book. Both are argued positions rather than neutral description, and this record names them as such rather than adopting the judgment as its own; but neither is a fringe reading. The manuscript that had taken a Safavid royal workshop over a decade to make, and that survived four centuries of documented and undocumented transit, was broken into market-tradeable units by its twentieth-century private owner, over the course of roughly twenty years, for financial reasons the sources consulted state without dispute.

This folio's own path since 1970

Set against that history, this specific object's own provenance since 1970 is unusually simple. It has remained in the Metropolitan Museum's collection continuously since the 1970 gift, under the accession number 1970.301.3, with no recorded sale, loan for sale, exchange, or deaccession in the sources consulted. Where the manuscript as a whole is a story of dismemberment and dispersal across dozens of institutions and private hands, this folio's individual history after 1970 is the story of an object that simply stayed where it was given, which is itself worth stating plainly rather than assuming: not every folio from this manuscript can say the same.

Where the rest of the manuscript went

The 78 illustrations given to the Met in 1970 and the individual lots sold at Christie's in 1976 and 1988 account for only part of the original 258. Further folios entered public and private collections across the following decades: ten folios are now in the Khalili Collection of Islamic Art, assembled by Nasser David Khalili, spanning several of the manuscript's named painters including Sultan Muhammad, Mir Sayyid Ali, Mir Musavvir, Aqa Mirak, and Mirza Ali, among them the manuscript's only dated folio. Other leaves are recorded in the Museum fur Islamische Kunst in Berlin, the Freer and Sackler Galleries of the Smithsonian Institution in Washington, the Los Angeles County Museum of Art, the David Collection in Copenhagen, the Museum of Fine Arts in Richmond, Virginia, the Museum of Islamic Art in Doha, the Reza Abbasi Museum in Tehran, and Harvard Art Museums, successor to the Fogg where Welch first worked on the material, alongside further leaves in private hands whose current whereabouts are not published. The Court of Gayumars, the manuscript's single most reproduced page, is in the Aga Khan Museum in Toronto; it is a different folio from the one described in this record and is discussed here only for that reason, to prevent the two being confused.

Vienna, 1994: the exchange for a de Kooning

Houghton died in 1990, still holding the manuscript's remaining text, binding, and roughly 118 paintings. His heirs first tried an intact sale again, reportedly seeking around 20 million dollars in 1991-1992, again without a buyer at that price. What followed instead was a negotiation, running from 1991 through 1994, conducted by the London dealer Oliver Hoare on behalf of the Houghton estate with officials of the Islamic Republic of Iran. Iran held, in the collection of the Tehran Museum of Contemporary Art, Willem de Kooning's 1953 abstract expressionist painting Woman III, acquired in 1977 under the pre-revolutionary government of Shah Mohammad Reza Pahlavi and Empress Farah Pahlavi as part of a major purchase of Western modern art; after the 1979 revolution the painting sat in a collection whose figurative Western holdings had become, at minimum, politically and religiously awkward for its new government to display.

The exchange itself took place on the tarmac of Vienna's airport on 28 July 1994: crates containing the remaining Shahnameh text, binding, and 118 paintings were handed to Iranian officials, and Woman III was handed to the American side, with the de Kooning reportedly moved quickly to a dealer in Switzerland out of concern the United States government might attempt to seize it against other disputed claims between the two countries. Farah Pahlavi has since criticized Hoare's role in the deal as disrupting the unity of her collection, a criticism aimed at the exchange itself rather than at the original 1970s breakup. The remaining manuscript material returned to Iran as a result; this record was not able to confirm a single specific current repository for it within Iran, beyond its delivery to the Iranian government via the Tehran Museum of Contemporary Art's negotiators, and that is recorded below as an open question rather than resolved with an assumption.

Reproduction: the book that survived the breakup on paper

Two publications, both tied directly to Houghton's ownership, mean the manuscript's dispersal is not the last word on it. Stuart Cary Welch's A King's Book of Kings appeared in 1972, two years after this folio and its companions entered the Metropolitan Museum, and used the newly accessible Met folios as its core illustrations while offering a full checklist of all 258 paintings, wherever they then stood. Welch and Martin Bernard Dickson followed in 1981 with The Houghton Shahnameh, a two-volume deluxe facsimile from Harvard University Press, limited to 750 numbered copies and priced at 2,000 dollars a set on publication, reproducing 20 plates in color collotype and hundreds more in monochrome across more than 800 combined pages of text and image.

That timing is worth noticing. The facsimile was compiled and published in the same decade Houghton was actively selling the physical folios at Christie's, which means the fullest scholarly record of the manuscript as a single object was assembled concurrently with, not after, its physical dismantling. A reader today can reconstruct the entire sequence of 258 paintings from Dickson and Welch's plates even though no single institution or owner holds the whole thing; the facsimile is, in a real sense, the only complete copy of the manuscript that still exists in one place.

What was not found

Two absences are worth recording rather than leaving silent. No source consulted for this record reports a published pigment analysis, X-radiography, or infrared examination of this specific folio; the technical literature on the manuscript's materials, where it exists, was not located in the sources available for this record, and the medium and materials given here rest on the Metropolitan Museum's own cataloguing rather than an independent technical study. Second, nothing in the sources consulted suggests a colonial-extraction dimension to this object's own path out of its country of production: it moved by diplomatic gift in 1568, by private sale among European and American collectors and dealers from the 1900s onward, and by museum gift in 1970, a chain documented, where it is documented at all, as commerce and diplomacy between willing parties rather than as seizure. That absence is itself worth stating plainly, since the manuscript's later history, the 1994 exchange, does raise real questions about an Iranian institution's post-revolutionary hand being forced by a de-luxe manuscript's market value against a painting its own government could no longer display; those questions belong to that later episode and are not evidence of anything at the point this folio itself changed hands.

What a folio from this manuscript sells for now

The individual-folio market this dispersal created has continued to set records in its own right. A folio attributed to Aqa Mirak, Faridun in the Guise of a Dragon Tests His Sons, from folio 42 of the same manuscript, sold at Sotheby's London in 2011 as the star lot of the Stuart Cary Welch collection sale for 7.4 million pounds, reported at roughly 12 million dollars, described at the time as a record for any work of Islamic art on paper. That is the strongest documented comparandum for what a single leaf from this specific manuscript can command decades after the breakup, and it stands in the same market lineage as the 1976 sale that first proved individual folios had a price at all. No sale of this folio itself, 1970.301.3, appears in the sources consulted; it has never left the Metropolitan Museum's collection to be tested against that market.