At the end of the chapter on “Technique, colours and paper” in his 1912 Miniature Painting and Painters of Persia, India and Turkey, F. R. Martin turns to the, for him, evidently irksome topic of forgeries of classical Persian paintings then appearing on the contemporary art market. Somewhat disingenuously (given what we now know of his own irregular activities), Martin proceeds to bemoan the likelihood that “this book will give the forgers plenty of models”. He then singles out, without naming names, a Persian artist who, sometime after 1908, had begun to work for the “dealers of Teheran, Constantinople and Paris”, and who, together with many pupils, every month executed hundreds of miniatures that were then passed off and collected as authentic works of art. Martin himself was once almost taken in by these forgeries, and was close to buying a copy, in the form of a black-and-white sketch, of a miniature in the Musée des Arts Décoratifs (MAD), Paris, reproduced in the second volume of his publication.Footnote ¹ The MAD miniature that Martin cites is, of course, none other than the celebrated composition of “Humay and Humayun in a garden” (Figure 1), acquired by the Paris museum in 1887 and to this day widely regarded as a masterpiece of Timurid painting.Footnote ²

Figure 1. “Humay and Humayun in a garden”, Iran, circa 1430. Source: Musée du Louvre, Paris (deposited from Musée des Arts Décoratifs, n°AD 3727). © Musée du Louvre, dist. RMN—Grand Palais/Raphaël Chipault.

It has long been recognised, and now reconfirmed as a result of rigorous art historical and technical investigations of its structure and composition undertaken in Paris, that the scene illustrates a poem by Khwaju Kirmani recounting the passion of Humay, a prince of Syria, for Humayun, daughter of the emperor of China.Footnote ³ Towards the beginning of the narrative, Humay has a vision in which he sees Humayun in a beautiful garden and falls under her spell. Three verses of Humay's dream rhapsody are included in the text panel on the upper section of the painting's landscape. As we now also know, thanks to recent in-depth study, these verses were written on the same paper substrate as the painting. A related point of interest is that the composition and verses once formed the recto of a manuscript folio, with the Khwaju Kirmani text continuing on the verso. The illuminated rubrics surrounding the painting on three sides, however, come from certain other sections of the Humay and Humayun story and presumably from the same Khwaju Kirmani manuscript to which the Paris miniature once belonged. The two vertical panels on the right side and the pair of horizontal ones immediately above the poetic verses are affixed directly onto the painted surface and so cover part of the composition. The bands of three horizontal panels at the top and bottom are pasted onto the cardboard sheet on which the painted folio is mounted.Footnote ⁴

To return to Martin and the conclusion to his querulous commentary on duplicitous artistic practices: “Forgers have bought all the photographs which were taken of the miniatures at the Munich exhibition which they will doubtless utilize in the preparation of more specimens of their fraudulent work.”Footnote ⁵ Here he is referring to the photography campaign that took place during the large, ground-breaking exhibition of Islamic art held in Munich in 1910, which included the Humay and Humayun painting from Paris. In addition to providing reproductions for newspaper and journal reviews at the time of the Munich show and for the three-volume post-exhibition catalogue published by Sarre and Martin in 1912, these photographs were sold at the exhibition itself.Footnote ⁶ While the idea behind the immediate availability and subsequent publication of such images was to promote the historical and aesthetic significance of Persian art and further its public reception, the unintended consequence may have been, as Martin predicted, to serve the particular interests of dealers and the artists in their employ. That those engaged during the early decades of the twentieth century in the production of, and trade in, “classical” Persian painting took advantage of the invaluable resource presented by the Munich photographs, as well as by the reproductions in Martin's book and other publications of the period, is suggested by a version of the iconic Humay and Humayun painting found today in Philadelphia (Figure 2).

Figure 2. “Humay and Humayun in a garden”, early twentieth century. Source: Free Library of Philadelphia, Rare Book Department, Lewis O.68, folio 185b of a Kulliyat of Sa'di.

In March 1927 the Philadelphia bibliophile and collector John Frederick Lewis acquired a deluxe manuscript of the Kulliyat of Sa'di, said to date from the mid-sixteenth century, at a New York auction of works from the Hagop Kevorkian collection.Footnote ⁷ The sales catalogue entry specified the volume's contents, including illuminated ‘unwans at the start of Sa'di's individual poems and 13 miniature paintings, with one identified as “Homay visiting Homayoun, the garden-scene of the palace”.Footnote ⁸ Interestingly, the entry did not reveal that much of the manuscript's illumination and all of its paintings were additions to the original production. This presumed oversight was partly corrected ten years later in the catalogue, compiled by Muhammed A. Simsar, of Lewis's “Oriental” manuscripts, by then in the Free Library of Philadelphia. Besides observing that the double-page pictorial frontispiece and finispiece in the Sa'di volume were painted later, Simsar observed in a footnote that the scene of Humay and Humayun was “similar in its style of portraiture to the well-known painting at the Musée des Arts Décoratifs, Paris”.Footnote ⁹ He does not, however, draw any inference as to the nature of that similarity or how it came about.

The general layout of the Free Library of Philadelphia manuscript is comparable to other volumes of Sa'di's Kulliyat, particularly those from Shiraz during the second half of the sixteenth century (Figure 3).Footnote ¹⁰ Each page features two written surfaces: one in the centre (or, actually, off-centre, as in traditional Islamic manuscripts) containing 13 horizontal lines of text, surrounded on three sides by another zone (more precisely, a marginal column), with 26 shorter lines altogether written on the diagonal in two complementary directions. The change in the lines’ direction is marked by a large illuminated equilateral triangle in the middle of the secondary zone; smaller right-angled illuminated triangles set off the inner or gutter corners. Both text areas are regularly punctuated with illuminated rubric panels. The start of each section of the Sa'di text is signalled by an illuminated title-piece, with facing pages of text in illuminated contour or cloud panels. Although the treatment of the illumination, with small, colourful buds and blossoms on a gold ground, is certainly comparable to that of sixteenth-century Shiraz, closer inspection reveals that much has been reworked. Indeed, the manuscript as a whole has experienced a great deal of decorative enhancement, including the addition of blue and gold floral sprays, pink page markers, and blue outer ruling lines in the margins, many of which are formed of replacement paper.

Figure 3. Kulliyat of Sa'di, Iran, Shiraz, second half of the sixteenth century, with early twentieth-century modifications. Source: Free Library of Philadelphia, Rare Book Department, Lewis O.68.

The extent of the manuscript's refurbishment is particularly apparent in the 13 miniatures, which are rendered in an awkward and heavy-handed manner (more manner than style) with a very thick application of pigment. The thickness of the paint layer was evidently intended to disguise the fact that various paintings were executed over written and illuminated pages. It is still possible, however, to make out parts of words underneath the painted surface (folio 211b) and even remnants of illumination incorporated into a composition (folio 142b) (Figure 4). In other instances, paintings were rendered on originally blank pages that fell between the end of one section of the Sa'di text and the beginning of another. Such is the case of the Humay and Humayun scene (folio 185b), which appears on the verso side of the page that ends the poet's Arabic qasidas (folio 185a) and facing the recto page that begins the Persian qasidas (folio 186a) (Figure 5).

Figure 4. “Mosque scene”, early twentieth century. Source: Free Library of Philadelphia, Rare Book Department, Lewis O.68, folio 142b of a Kulliyat of Sa'di.

Figure 5. “Humay and Humayun in a garden” and the beginning of Sa'di's qasidas, early twentieth century. Source: Free Library of Philadelphia, Rare Book Department, Lewis O.68, folios 185b–186a.

While there is no doubt that the Philadelphia painting is a copy of some form of reproduction of the Humay and Humayun painting in Paris that was available before the Kevorkian sale of 1927—either as a published reproduction or a photographic print from the 1910 Munich exhibition—it seems doubtful that the copyist was trying to make an exact duplicate. In fact, the Philadelphia picture deviates in a variety of noticeable ways, both large and small, from the Paris original. The most striking difference is in the overall stiffness and blockiness of the four figures and the elements of their landscape setting. Whereas the Paris figures, flowering plants, and trees epitomise the grace, charm, and delicacy of the courtly Timurid style, their Philadelphia counterparts are stolid, spiritless, and far from the “paradise on earth” envisioned in Humay's dream. Similarly, the Philadelphia palette is flat and dull, and lacks the translucency and sheen of the Paris palette.Footnote ¹¹ The overall effect of the Paris painting is at once intense and joyful; by contrast, the Philadelphia version is devoid of life. As for compositional particulars: a section of the garden fence or balustrade panel at the left of the Paris scene has been removed, resulting in a compression of the painting's relative width, and the gold panels under the scalloped arches have been replaced by a tile pattern in light brown on a white ground. The rectangular fountain or water basin at the left, which in the Paris painting debouches through two gold faucets into the upper stream, has lost its original function and now looks something like a garden planter. The stream itself lacks the eddies and flows of the Paris original, while its mate in the bottom edge has been removed altogether. Even more evident is the disappearance of the three Khwaju Kirmani verses from the picture plane; in the Philadelphia version these have been replaced by two panels of light grey tiles and by the extended branches of the flowering tree that grows “behind” the text block in the Paris painting.Footnote ¹² Furthermore, the number of illuminated rubrics pasted onto and around the Timurid miniature, and painted as part of the surface of the Philadelphia scene, have been pared down, with two eliminated from the top centre and top left and the vertical one from the lower right side.

The central rubric at the bottom provides two other—perhaps the most tell-tale—indicators that the artist responsible for the Philadelphia painting was trying to mask any connection with the Paris original. As is today well documented, the Musée des Arts Décoratifs purchased its beautiful Humay and Humayun composition from Nicolas-Alexis Jaroszinski, a naturalised French citizen of Polish origin, who served as an interpreter at the French embassy in Constantinople/Istanbul from 1882–1887.Footnote ¹³ During that time, Jaroszinski also occasionally served as an intermediary between Istanbul notables and the museum. The Polish-French official returned to Paris in failing health and on 31 May 1887 sold his miniature to the museum. The painting was then marked with an inventory or accession number “A. N^o. 3727” in the left-hand corner of the central rubric, with the “A” standing for Achat or purchase. This mark was clearly visible in all the early twentieth-century reproductions of the painting and remains so today. Its omission from the Philadelphia copy was surely intentional, so as not to reveal the composition as a copy of a work in a recognisable museum collection. The Philadelphia painting also lacks a second sign of institutional ownership found on the Paris miniature, namely the small oval seal enclosing the letters A.D., meaning Arts Décoratifs, impressed just above the corner of the lower-left illuminated rubric and partly over a flowering plant with orange blossoms.Footnote ¹⁴

Such efforts at obfuscation were not unique, however, to the Philadelphia copy of the Paris painting. When Jaroszinki sold the painting in Paris, it already had been mounted and “framed” with illuminated rubrics from a Khwaju Kirmani manuscript—in other words, transformed from a book illustration into something resembling an album painting. Jaroszinski's own source has yet to be verified, but given that he worked in a diplomatic capacity in Istanbul, it is fair to assume that the painting came from the Ottoman court collections. Two decades later, when F. R. Martin served at the Swedish embassy in Istanbul, he too took advantage of his diplomatic status and the evidently favourable circumstances at court to “mine” the Ottoman holdings. One of his acquisitions was the “Portrait of Sultan Husayn Mirza” ascribed to Bihzad. Interestingly, the mount on which this painting, now known to have come from the Bahram Mirza album (Topkapi Palace Museum, H. 2154), is affixed includes an illuminated rubric from the same Khwaju Kirmani manuscript from which the Humay and Humayun painting and its pasted-on illuminations were removed.Footnote ¹⁵ It is tempting to imagine that Martin got his hands on the very same—and by then mutilated—Khwaju Kirmani manuscript to which the Humay and Humayun miniature once belonged and cut out another rubric to decorate the mount for Sultan Husayn Mirza's portrait.

While a notable bibliophile and avid collector of Islamic codices, John Frederick Lewis did not necessarily follow the scholarly literature in the Islamic manuscript field, and it seems unlikely that he would have been familiar with the Timurid painting of Humay and Humayun, as reproduced in the publications by Migeon (1907), Martin (1912), Sarre and Martin (1912), and Kühnel (1922).Footnote ¹⁶ On the other hand, Hagop Kevorkian, the previous owner of the Sa'di manuscript in Philadelphia, certainly would have known of those early reproductions, and doubtless also the original painting itself as exhibited and photographed in 1910 since he himself lent it to the Munich show. It is even possible, following Martin's derogatory comment about dealers in Miniature Painting and Painters, that Kevorkian commissioned the paintings for the Kulliyat volume that he sold in New York in 1927, on the doubtlessly correct assumption that a potential American buyer, such as Lewis, might be able to distinguish between a silk purse and a sow's ear, but not between classical and modern production in Persian painting.Footnote ¹⁷