Making the Invisible Visible: Conservation and Islamic Art

Beth Edelstein
Associate Conservator
The Sherman Fairchild Center for Objects Conservation

Damascus Room, dated A.H. 1119/A.D. 1707. Syria, Damascus. The Metropolitan Museum of Art, New York, Gift of The Hagop Kevorkian Fund, 1970 (1970.170)

As part of the conservation of the Damascus Room, conservators and conservation scientists conducted an in-depth study of the room with an eye toward better understanding its original appearance and layout. This project made a significant contribution to the study of Syrian reception rooms worldwide. Similar interiors can be found in museum collections, universities, in private hands, and in their original locations in homes in Damascus, Aleppo, and Hama.

The results of the study were shared with the conservation community at the International Institute for Conservation's biannual conference in Istanbul in September 2010, and at the Architectural Paint Research Conference in Lincoln, U.K., in August 2010. Related articles and poster publications are available below, in addition to related publications and links to information about other Syrian interiors.

Publications

Baumeister, M., Edelstein, B., Rizzo, A. et al. 2010. "A splendid welcome to the house of praises, glorious deeds and magnanimity." In Conservation and the Eastern Mediterranean. International Institute for Conservation of Historic and Artistic Works (IIC) Contributions to the Istanbul Congress, Rozeik, C., Roy, A. and Saunders, D. (eds), 20–24 September 2010. London: IIC.126–133.

Schultz, J., Arslanoglu, J., and Petersen, K. 2010. "The use of ELISA for the identification of proteinaceous binding media from an eighteenth-century Damascene reception room." Poster presentation in Conservation and the Eastern Mediterranean. International Institute for Conservation of Historic and Artistic Works (IIC) Contributions to the Istanbul Congress, 20–24 September 2010. Rozeik, C., Roy, A. and Saunders, D. (eds) London: IIC. 269.

Fair, L., Rizzo, A., and Edelstein, B. 2010. "Painting techniques in Ottoman interiors." Poster presentation in Conservation and the Eastern Mediterranean. International Institute for Conservation of Historic and Artistic Works (IIC) Contributions to the Istanbul Congress, 20–24 September 2010. Rozeik, C., Roy, A. and Saunders, D. (eds) London: IIC.

Scharrahs, Anke. Damascene Ajami Interiors: Forgotten Jewels of Interior Design. London: Archetype books, 2013

Related Links

The Damascus Room at Shangri La, the Doris Duke Foundation for Islamic Art, Honolulu, Hawaii
http://shangrilahawaii.org/Islamic-Art-Collection/Collection-Highlights/Late-Ottoman-Syrian-interiors-and-furnishings/Damascus-Room/

'Ajami Wood Paneling in the Damascus Room at Shangri La. Presented by Anke Scharrahs
http://vimeo.com/47620230

The Syria-Lebanon Room at the University of Pittsburgh Cathedral of Learning
http://www.pitt.edu/~natrooms/syria-lebanon/syri.html#

The Dresden Damascus Room
http://www.skd.museum/en/museums-institutions/japanisches-palais/museum-fuer-voelkerkunde-dresden/the-dresden-damascus-room/index.html

Related Videos

Conserving The Damascus Room

True Colors: The Damascus Room

Discoveries Symposium Part 4

Jean-Francois de Laperouse
Conservator
The Sherman Fairchild Center for Objects Conservation

Bowl. Seljuq period (1040–1196), 12th–13th century. Attributed to Rayy, Iran. Stonepaste; polychrome in-glaze and overglaze painted and gilded on an opaque white glaze (mina'i). H. 3 1/2 in. (8.9 cm), Diam. 8 1/4 in. (21 cm). The Metropolitan Museum of Art, New York, Rogers Fund, 1913 (13.93.1). See a slideshow of mina'i vessels that are on view or in storage.

At the start of the reinstallation project, conservators surveyed the entire collection with curators to establish examination and treatment priorities. The condition of the ceramics, including the mina'i ware, was of particular concern as it was known that many had been restored in the past but it was not clear how much of their original fabric survived under discolored restoration paint. Produced in the Kashan region of Iran in the late twelfth–early thirteenth century, mina'i vessels were covered with an opaque lead-tin oxide glaze—either white or tinted light blue with copper salts—that was meant to imitate Chinese porcelain. Their surface decoration was executed using both inglaze colors—pigments that suffused into the upper surface glaze in a first firing—and overglaze enamels consisting of pigments mixed with glass frit that melted and fused to the glaze in a second firing. Unfortunately, the ceramic body used—a mixture of crushed quartz, glass frit and white clay—was not as durable as porcelain, and almost all examples of mina'i ware have been incompletely retrieved in fragmentary form. Previous restorers filled losses with plaster, fragments from other old vessels or newly fired ceramic shards which they then disguised with paint.

Ultraviolet (UV) light can be an extremely useful tool in identifying modern restoration paint on ceramics. The binders in the paint—the drying oils or organic resins that keep the pigment particles together and affix them onto a surface—will absorb UV light and reflect a visible fluorescence that will contrast with the unpainted areas. Since paint was often applied over original surfaces to hide break lines and blend large restorations into surrounding areas, the UV reflectance images by themselves did not provide a complete picture of an object's condition. Comparing UV reflectance images with radiographs taken of each vessel, however, allowed us to determine accurately the exact extent of losses. (Records were kept of all radiographic exposures to ensure that any possible effect on future thermoluminescence dating could be accounted for in the future.) Using the information obtained, photographs could be edited digitally to illustrate various treatment scenarios before treating the objects themselves.

Through discussions with the Islamic curators, it became clear that our primary concern was to reveal original surfaces as much as possible and to remove extraneous shards. In some cases, old restorations were preserved if they were not misleading and were confined to areas of loss, as some of these restorations were almost one hundred years old and had acquired their own historic value. When the repainting of losses was required, the restoration of original motifs such as repeated designs or framing was done occasionally so that losses would not be too visually obtrusive.

See a slideshow of mina'i vessels that are on view or in storage.

Additional Reading

de Lapérouse, J-F, K. Stamm and V. Parry, "Re-examination and Treatment of Mina'i Ceramics at The Metropolitan Museum of Art," in Pilosi, Lisa, ed., Glass and Ceramics Conservation 2007: Preprints of the Interim Meeting of the ICOM-CC Working Group, August 27–30, 2007, Nova Gorica, Slovenia (Nova Gorica, Goriski Muzej Kromberk, 2007), pp. 112–119.

Koss, K., B. McCarthy, E. S. Chase and D. Smith, "Analysis of Persian Painted Minai Ware." In Scientific Research on historic Asian Ceramics: Proceedings of the fourth Forbes Symposium at the Freer Gallery of Art, Archetype Publications Ltd (2009), pp. 33–47.

Pease, M., "Two Bowls in One." The Metropolitan Museum of Art Bulletin, New Series, Vol. 16, No. 8 (April 1958), pp. 236–240.

Beth Edelstein
Associate Conservator
The Sherman Fairchild Center for Objects Conservation

Courtyard and fountain, 'Attarin Madrasa, Fez, Morocco, 1323–25

Panel with Cusped Arches, 14th century. Morocco. Wood (cedar); carved and painted; H. 19 in. (48.3 cm) W. 121 in. (307.3 cm) D. 2 3/4 in. (7cm) Wt. including crate 247 lbs. (112 kg) crate maybe half of this weight. The Metropolitan Museum of Art, New York, Mr. and Mrs. Isaac D. Fletcher Collection, Bequest of Isaac D. Fletcher and Rogers Fund, by exchange, 1985 (1985.241)

This frieze was likely part of the decoration of an outdoor courtyard, and shows the remains of mortise and tenon joints as well as large hand-wrought iron nails used to attach it to adjacent panels. The carved decoration, which includes the word yumn, meaning "good luck," has been repainted at least once with a bright red, yellow, blue and white color scheme. These later paint layers were not well adhered to the surface, and the colors were obscured by a haze of dust and grime.

Before treatment(left), After treatment (right)

Detail of surface before treatment

Humidifying the thick curled paint layer with an ultrasonic mister allowed the paint to be flattened and readhered to the underlying surface. A layer of thin mulberry-fiber tissue was temporarily laid over the surface to keep the paint flat as it dried.

The original design, evident in some areas beneath the later flaking paint layer, included black and white outlines and dots on a red (shown here) and blue background. Other areas have a yellow underlayer with an orange-toned glaze.

Additional Reading

Terrasse, Charles, Médersas du Maroc. Paris: A. Morancé, 1928.

Paccard, Andre. Traditional Islamic craft in Moroccan architecture; translated from the French by Mary Guggenheim. Saint-Jorioz, France : Éditions Atelier 74, 1980.

Fred Sager
Senior Conservation Preparator
The Sherman Fairchild Center for Objects Conservation

Dagger with Zoomorphic Hilt. Islamic, Adil Shahi period (1490–1686), second half 16th century. Hilt: copper; cast, chased, gilded, and inlaid with rubies. Blade: steel; forged; L. 15 5/8 in. (39.6 cm). Attributed to: Bijapur or Golconda, Deccan, India. The Metropolitan Museum of Art, New York, Purchase, Lila Acheson Wallace Gift, 2011

Mounting objects so that they can be displayed safely is an important part of preventative conservation. Preparators in the Museum's Objects Conservation Department fabricated hundreds of mounts for the reinstallation of the Islamic galleries that by design are intended to be as invisible as possible. They worked with conservators, curators and exhibition designers to ensure that their individually crafted brass mounts would hold artwork securely at the desired viewing angles. Larger and heavier objects are supported by steel mounts that were designed and produced by the staff of the Museum's Metal Shop. All mounts were padded to prevent damage to fragile surfaces and painted to match the color and texture of the artwork as closely as possible.

Vicki Parry
Associate Conservator
The Sherman Fairchild Center for Objects Conservation

From 1935 until just after the Second World War, The Metropolitan Museum of Art conducted archaeological excavations at the site of the early Islamic city of Nishapur in Iran. In accordance with Iranian antiquity laws and excavation agreements, half of the excavated finds went to the National Museum of Iran in Tehran and the remaining material came to this Museum.

Although considerable efforts were made on site and later in the Museum in the 1970s to prepare this material for exhibition and publication, the use of unstable adhesives—primarily cellulose nitrate on the ceramics, architectural stucco and wall paintings and polyester resin on the glass—had made much of the collection physically unstable by the time the previous galleries were de-installed. In addition, the yellowing and darkening of the adhesives and surface consolidants previously used made it difficult to see original colors and designs. Also of concern were the elevated levels of potentially damaging soluble salts deposited during burial that had never been adequately removed. Finally, many ceramic and glass vessels had excessively large fills, which were sometimes misleading in shape and which limited access to break edges for compositional study.

Left: Ceramic restoration workroom on site at the Nishapur excavations; Right: Conservation of the collection today

In 2008, the Institute of Museum and Library Services awarded a matching grant to stabilize the excavated material from Nishapur. Thanks to this funding, the poor condition of most of the glass, ceramics, and stucco finds could be addressed while the entire collection was in storage. Prior to treatment, the condition of objects was documented with digital images taken in natural and ultraviolet (UV) light, the latter of which was useful in revealing the full extent of previous restorations. Modern paint and old adhesives were softened and removed by swabbing with solvents, allowing the shards to be freed of previous fills. Desalination of the ceramics in distilled water not only reduced soluble salt levels but also provided the least invasive and most effective way of removing excess restoration plaster from porous surfaces.

Natural (left) and UV (right) images of bowl (39.40.147). In the image on the right the old restoration paint fluoresces brightly.

Once ceramic fragments were cleaned and stabilized, they were rejoined using stable and reversible acrylic resins. In general, losses were not filled unless they were required to allow the vessels to be handled safely. All fills were done using reagent grade gypsum plaster or glass micro-balloons in acrylic resin.

Images of bowl (38.40.148) before treatment (left) and after treatment (right)

Glass objects, on the other hand, required transparent fills made with an epoxy resin that presented several technical challenges. This project benefited enormously from the expertise of Goradz Lemajič of the National Museum of Slovenia, who has pioneered the use of molded PVC film to create precisely contoured molds into which clear epoxy resin is injected. The technique of Mr. Lemajič, who served as a consultant to this project, has been further developed in the Museum for use on glass objects in the entire collection.

Left: Preparation for filling loss in in glass vessel (40.170.61); Right: Injecting epoxy resin in an area of loss in glass vessel (40.170.61)

Finally, in addition to ceramics and glass, carved stucco architectural panels and wall paintings required urgent treatment due to flaking and the discoloration of old adhesives and consolidants.

Ongoing Study of the Nishapur Collection

The development of new analytical techniques since the Nishapur material was originally excavated provides an opportunity to glean more information from this important archaeological collection. The composition of the glass from the site is being analyzed by Mark Wypyski, Research Scientist in the Department of Scientific Research, and preliminary research has been undertaken toward future technical studies of the ceramics and wall paintings.

Additional Reading

Kröger, J., Nishapur: Glass of the Early Islamic Period (New York, Metropolitan Museum of Art, 1995).

Wilkinson, C. K., Nishapur: Pottery of the Early Islamic Period (New York: Metropolitan Museum of Art, 1973).

———Nishapur: Some Early Islamic Buildings and their Decoration (New York: Metropolitan Museum of Art, 1986)

Beth Edelstein
Associate Conservator
The Sherman Fairchild Center for Objects Conservation

Front (left) and reverse (right) views of a Pair of Carved Doors, 1466. The Metropolitan Museum of Art, New York, Rogers Fund, 1935 (35.127a, b).

This pair of carved doors with inset panels bears a date of 1466 on one of the small rectangular panels at the bottom, and the name of the maker on the vertical bar. During conservation examination and treatment, three handwritten inscriptions were discovered on the reverse, which became readable only with near-infrared reflectography (IRR). IRR photography capitalizes on the sensitivity of digital cameras to wavelengths in the infrared spectrum, which are absorbed or reflected by different materials. Inks or other black materials commonly used for writing or underdrawing absorb infrared, allowing their visibility even when faded or obscured by surface layers. Initial translation of the longest inscription, written in Persian, suggests that it describes the difficulty in completing a task—perhaps referring to the creation of the doors themselves, or of their larger setting.

Detail of inscription in visible light (left) and with infrared reflectography (right)

Detail of inscription in visible light (left), and with infrared reflectography (right). Below: Detail of longest inscription, referring to the difficulty of completing a task.

Sarah McGregor
Associate Conservator
The Sherman Fairchild Center for Objects Conservation

Fig. 1. Digital photographs taken before treatment

This ewer in the form of a rooster was most likely made in Kashan in Iran in the thirteenth century using stoneware, a composite of crushed quartz, glass frit, and white clay (fig. 1). While the body and neck were formed on a potter's wheel, the head was made in a mold and the handle was modeled separately. After drying, the head and the handle were luted onto the body using a clay slip. A blue underglaze paint was applied to delineate the features on the head and decorate the body with vertical stripes that alternate with the tan color of the stoneware body. A clear alkaline glaze was then applied over the entire surface before the ewer was fired. The fine craquelure seen in this glaze may have occurred during cooling when the glaze and the ceramic body did not shrink at equal rates resulting in what is known as a poor "fit."

Fig. 2. Digital photograph taken with ultraviolet light. The whitish blue is the fluorescence from the adhesive and the flat pale yellow fluorescence is from the paint and plaster fills.

Prior to its acquisition in 1919, this ewer had been repaired after sustaining damage to the body, handle and feathers of the rooster's crown. In addition, the glaze had fallen off in isolated areas. At that time, the cracks in the glaze had been consolidated with an unidentified adhesive and the areas of glaze loss, as well as the small losses in the crown, had been filled and disguised with paint that often extended over original surfaces as revealed when viewing the surface under ultraviolet light (fig.2).

Fig. 3. A digital photograph of the ewer before treatment. The extraneous fragment is located between the yellow arrows.

While the adhesive joins in the body were well aligned and appeared to be stable, the joins in the handle were not. The middle section of the handle is lost and had been replaced with a poorly fitting handle fragment from a similar vessel forming a pastiche (fig. 3). There were marked differences between the color, texture, and thickness of this extraneous fragment and the rest of the ewer.

Fig. 4. The mock-up of the new handle made with plasticine clay.

In consultation with a curator, the decision was made to replace the handle fragment with a new fill that would more accurately represent the more delicate proportions of this ewer. It should be noted that the handle is meant to represent a rooster's tail consisting of one or more feathers with curled ends. After looking at comparable examples in other collections it was decided to make a single feather strand. A mockup of the new handle was made and molded to produce a plaster cast (fig.4).

Fig. 5. Digital images taken after treatment.

This new plaster handle was joined in place with a stable acrylic resin and painted so that it matches the color of the ewer but can be discerned upon close inspection (fig. 5). The old restoration paint was removed and the losses also were painted with acrylic emulsion paints.

Additional Reading

Keblow Bernsted, Annie-Marie, Early Islamic Pottery—Materials and Techniques. Archetype Publications Ltd., 2003.

Beth Edelstein
Associate Conservator
The Sherman Fairchild Center for Objects Conservation

Ceiling, 16th century. Spain. The Metropolitan Museum of Art, New York, Gift of The Hearst Foundation, 1956 (56.234.35.2)

During the conservation of this monumental wooden ceiling, conservators discovered that it had been considerably altered before entering the collection, possibly in an early twentieth-century restoration for William Randolph Hearst. The ceiling had been enlarged by approximately sixty percent, and its original five-faceted profile was converted to the current three-faceted shape. The wooden boards added at this time were distressed to mimic the flaking paint of the original areas, and new elements of the geometric design were incorporated, resulting in a geometric configuration that is no longer mathematically sound. Conservators also discovered that the carved medallions in the blue fields were originally gilded, but had been painted red during the restoration. The conservation study also analyzed the materials, joinery, and decorative techniques of the ceiling, and found closely related examples from the Castile-Leon region of Spain, shedding new light on the ceiling's provenance. The results of the study were published in the Journal of Architectural Conservation in November 2010.

Shtrum, B., Brussat, M., Garcia, M., Hayes T., Massaux, S. "The Metropolitan Museum of Art's 'Spanish Ceiling' Project: Interpretation and Conservation." In Journal of Architectural Conservation Vol 16, No. 3 (November 2010), 29–50.

Text and fiber microscopy by Florica Zaharia
Conservator in Charge, Department of Textile Conservation

Tapestry was the first type of weaving that allowed for the execution of multiple motifs by an interchange of colors in the weft direction. The techniques used to join the wefts and delimit motifs are slit-tapestry, dovetailing, and hatching, as well as single and double interlocking. Tapestry weaving reached its apogee in western Europe in the fifteenth and sixteenth centuries, when monumental wall hangings competed for prominence with paintings and frescos.

In the Islamic world, the tapestry method was used to weave fine carpets, hangings, and costumes; particularly favored were dovetailing, double interlocking, and slit-tapestry techniques (figs. 1, 3, 4). While hatching was crucial to the depiction of light and shadow in the illusionistic imagery of European tapestries, the stylized and geometric patterns of Islamic examples were achieved through color gradations. Additionally, in contrast to the more common plain-weave technique, a unique twill double-interlocking weave flourished in India and was used to create Kashmiri shawls and saris (fig. 2).

The materials used in Islamic tapestries, including silk, fine wool, cotton, and metal thread wound around a silk core (figs. 5–8), were exceptionally fine, allowing for the creation of the most intricate and delicate tapestries known today. Examples of all these techniques and materials appear in the exhibition galleries.

Fig. 1a, b. Plain weave, double- interlocking, front (left) and back (right). Photographs by Florica Zaharia

Fig. 2a, b. Plain weave, dovetailing, front (left) and back (right). Photographs by Florica Zaharia 

Fig. 3a, b. Plain weave, slit, front (left) and back (right). Photographs by Florica Zaharia

Fig. 4 a, b. Twill weave, double-interlocking, front (left) and back (right). Photographs by Florica Zaharia

Left: Fig. 5. Silk fiber, longitudinal view (400x); Right: Fig. 6. Fine wool fiber, longitudinal view (400x). Photographs by Florica Zaharia

Left: Fig. 7. Cotton fiber, longitudinal view (400x). Photograph by Florica Zaharia; Right: Fig. 8. Metal-wrapped silk thread, longitudinal view (200x). Photograph by Janina Poskrobko

Sarah McGregor
Associate Conservator
The Sherman Fairchild Center for Objects Conservation

Fig. 1. Tile Panel, ca. 1430. Syria, Damascus. The Metropolitan Museum of Art, New York, Gift of Professor Maan Z. Madina and Dr. Marilyn Jenkins-Madina, 2009 (2009.59.2a–p)

Conservators, curators, and museum preparators often work together to devise creative solutions for the safe and aesthetically pleasing display of fragmentary objects. This collaboration was critical to the successful treatment and mounting of a large panel of tiles from Syria dating to the fifteenth century. This panel, measuring approximately four square feet, originally consisted of twenty-five tiles, of which the Museum has eighteen (fig. 1). These tiles, which would have adorned the interior of a building, were made using stonepaste, a composite of crushed quartz, glass frit, and white clay that was pressed into a frame to obtain equally sized tiles. The raised molding found on the tiles was shaped separately by hand and attached to the surface using a clay slip (fig. 2).

Fig. 2. Above: The break edge of a tile fragment, showing the thin, horizontal line between the molding and the base where the molding was not sufficiently bonded to the base; Below: The side break edge of a tile fragment that is stained from the animal glue that was used in a previous restoration. The unglazed surface can be seen on the top of this tile, where the molding had once been attached with slip. Raised molding is often vulnerable to loss, especially if the slip has not bonded the molding to the base sufficiently.

A grid of incised lines, which can still be seen within the octofoil in the center of the panel, was laid out as a guide for painting after the initial bisque-firing (fig. 3). The black outlines of the floral and the geometric designs were then painted, followed by the cobalt blue and turquoise glazes, all of which were coated with a transparent glaze before refiring.

Fig. 3. The yellow arrows point to the incised lines that were made in the stoneware before the tile was glazed and fired.

The tiles had suffered varying degrees of damage over the years, as well as poorly executed restorations that had become discolored with time (fig. 4). Many of the tiles had chipped or missing corners and some had sections of the missing molding that had been replaced with plaster and/or ceramic molding taken from similar tiles. After a complete examination, it was decided to remove all previous restorations. In a few cases, the extraneous molding fragments were re-adhered after removal and cleaning if their shape and size closely matched what was lost. Otherwise, the extraneous fragments were removed, labeled, and returned to the Department of Islamic Art for separate storage.

Fig. 4. One of the tiles before treatment

New plaster fills were made separately and adhered in place using a stable and reversible acrylic resin. All fills were painted to match surrounding decoration as closely as possible (fig. 5)

Fig. 5. The same tile from fig. 4, after treatment.

After all the tiles had been treated, one question remained: how to display this fragmentary tile panel in such a way that visitors could appreciate its original visual impact. It was agreed that the missing areas should be represented by a drawing on the backing board, to which the tiles would be mounted. Brian Cha, from the Museum's Graphic Design Department, took careful measurements and created digital images of the tiles. Using a CAD (computer-aided design) program, he produced a full-scale drawing on which the existing tiles were placed to obtain a correct alignment between the drawing and the tiles. Once adjustments were made, a large sheet of tracing paper was placed over the drawing and the missing areas of the design were traced by hand. The tracing was then scanned and converted into a form that would allow it to be silkscreened onto a lightweight but strong honeycombed panel.

The next step was to mount the tiles securely to the panel with the faces and edges of the tiles flush and tightly together.

Fig. 6. Fred Sager, Senior Preparator

Being hand made, the tile backs are not perfectly flat, and slight variations in thickness exist. The original craftsmen had a yielding bed of mortar to accommodate these inherent variations. Working without mortar, Fred Sager, our Senior Museum Preparator, mounted the tiles by custom designing brass clips for each tile (fig. 6). In several areas, brass washers were stacked onto the clips to fill gaps between the tile back and the honeycomb board (fig. 7). All of this work took place over a sheet of clear Mylar to prevent any scratching to the silkscreened drawing. After all of the clips had been made, the Mylar sheet was removed, the clips were attached to the panel with screws, and the tiles were gently slid into place. The visible parts of the clips then were painted with acrylic emulsion to match the adjacent tile surfaces.

Fig. 7. Side view of the tile panel. The white arrow points to a clip that has a stack of brass washers supporting the unevenness of the back of the tile.

Additional Reading

Keblow Bernsted, Anne-Marie. Early Islamic Pottery Materials & Techniques, Archetype Publications, 2003.

Degeorge, George and Yves Porter. The Art of the Islamic Tile, 2002. Published originally as L'art de la ceramique dans l'architecture musulmane by the same authors in 2001 by Flammarion.

Janina Poskrobko
Conservator, Department of Textile Conservation

The manner in which textiles are displayed is determined by the condition of a textile as well as the materials and techniques used in its fabrication. Some of the methods most frequently used by the Museum's textile conservators to ensure the collection's long-term preservation are represented in the exhibition.

A well-preserved textile, such as the fragment from the dalmatic of San Valerius (46.156.10), can be displayed in a vitrine on a flat or slanted, low-angled support without any mounting as long as the ambient environmental conditions and lighting levels are suitably controlled.

Fragile textiles are often pressure mounted, which involves positioning a textile on a solid cushioned support without any stitches and holding it in place with light pressure from a Plexiglas sheet or box (fig. 1). This type of sealed mounting, used for the double-faced Kashan tapestry (43.84), allows for complete reversibility and environmental control. In addition, pressure-mounted textiles can be hung vertically.

Large textiles and those in varying states of preservation can be partially displayed. For example, one end of the chadar (41.19) is displayed in the exhibition on a slanted mount after consolidation to a support fabric, while the rest remains rolled on a suspended tube.

Finally, the safe display of a fragile costume such as the child's coat (1983.494.10) requires the insertion of archival materials to support its weight and restore its original shape. In this case, couching stitches secured areas of loss to an underlying support fabric of compatible color (fig. 2 a–c).

Fig. 1. Cross-section of a single pressure mount. Diagram by Florica Zaharia

Fig. 2. Bottom edge of a child's coat, before, during and after treatment. Photographs by Kisook Suh

Yael Rosenfield
Assistant Conservator, Department of Textile Conservation

Dye analysis provided by Nobuko Shibayama
Associate Research Scientist, Department of Scientific Research

Fiber Microscopy by Florica Zaharia
Conservator in Charge, Department of Textile Conservation

The Emperor's Carpet, second half 16th century. Iran, Safavid period (1501–1722). The Metropolitan Museum of Art, New York, Rogers Fund, 1943 (43.121.1)

Weave Structure

The carpet's weave structure is composed of silk warps intersected by a repeating pattern of one row of wool knots and three passes of silk wefts. In tightly woven carpets, warps are often dense, causing them to be pushed into two levels, with every other warp "depressed." Using silk for the foundation creates a finer weave and therefore a higher knot count; there are approximately 5,425 knots per square decimeter (equivalent to a 4 x 4 in. square). Dense warps and a high knot count allow for more intricate and accurate rendering of motifs, clearly demonstrated by the extremely fine design of this carpet.

Above: Sketch of the carpet weave structure. Drawing by Janina Poskrobko

Left: Asymmetrically knotted wool pile (Persian or Sehna knots), knots open to the left (front view); Right: Back view showing partially depressed silk warps and a repeating pattern of three silk weft rows and a row of dyed wool knots. Photographs by Yael Rosenfield

When several weavers work simultaneously on a wide carpet, the areas where discontinuous weft threads meet are marked by a series of diagonally arranged points called "lazy lines." These lines are found in the top two corners of the carpet.

Left: Lazy line at the top proper-right corner (front view); Right: Lazy lines are more visible on the back of a carpet. Photographs by Yael Rosenfield

Fibers and Dyes

The foundation of the carpet is composed of silk, a precious and rare fiber characteristic of carpets commissioned by rulers; the pile is a fine quality sheep's wool. The carpet's approximately sixteen colors were all derived from natural plant and insect materials: redwood, madder, weld, indigo (plant based), and lac (insect based). Dye analysis was performed by the Department of Scientific Research using high performance liquid chromatography (HPLC).

Left: Silk fiber used for warp and weft, fiber used for warp and longitudinal view (400x); Right: Silk weft, cross-sectional view (1000x). Photographs by Florica Zaharia

Left: Fine wool fiber used for pile, longitudinal view (400x); Right: Fine wool fiber used for pile, cross-sectional view (1000x). Photographs by Florica Zaharia

Conservation Treatment

To strengthen fragile areas, the carpet was restored in the early twentieth century, before its acquisition by the Museum in 1943. More than seven hundred patches had been applied to the reverse of the carpet, often sewn in layers, indicating they were applied over several time periods (see image below right). Many losses also had been mended with thick threads and too-tight stitches, causing distortion of the carpet's alignment. To document the history of the carpet and as a part of the Museum's conservation process, the patches were numbered and photographed before removal. All previous restoration also was removed, which allowed the carpet to settle and resume its original shape.

Overall view of the front (left) and back (right) before conservation. Photographs by The Photograph Studio

A support backing was created in order to strengthen the carpet and fill in areas of loss. The wool backing was custom dyed in four colors and its layout accurately drafted so that its main color areas would correspond to those of the carpet. Some "windows" were left uncovered; these were chosen because the carpet was in fairly good condition and would provide access to the reverse of the carpet for future examination. All losses were consolidated to the backing fabric using embroidery floss and a laid couching stitch.

New wool support backing, back view. Photograph by Yael Rosenfield

Left: Bottom proper-right corner after removal of old restoration, before conservation treatment; Right: Bottom proper-right corner after conservation, with couching stitches visible. Photographs by Yael Rosenfield

Transport, Display, Storage

When being moved or stored, the carpet is rolled together with archival materials on a large-diameter tube. It is rolled with its front facing outward, in the direction of the pile. During its six-month display rotation, it is exhibited on a custom-built adjustable platform. Between rotations, it is stored in the Museum's Ratti Textile Center on a suspended tube so as not to bear its own weight.

Installation of the carpet on its custom-made platform. Photograph by Yael Rosenfield

Additional Reading

Dimand, M.S. Oriental Rugs in the Metropolitan Museum of Art. The Metropolitan Museum of Art, 1973.

A.F. Kendrick, A. Upham Pope, W.G. Thomson. The Emperor's Carpet and Two Others, London and Persia: Cardinal & Harford, 1928.

Klose, Christine. "Imperial Puzzle: Sixteenth-Century Persian Spiral Vine Carpets with Animals," in Hali: Carpet, Textile and Islamic Art, Issue 170, Winter 201.

Volker, Angela. Oriental Knotted Carpets in MAK, Vienna, 2001.

Daniel Walker and Florica Zaharia. Masterpieces from the Department of Islamic Art in the Metropolitan Museum of Art, catalog 181, The Metropolitan Museum of Art, 2011.

Janina Poskrobko
Conservator, Department of Textile Conservation

Dye analysis provided by Nobuko Shibayama
Associate Research Scientist, Department of Scientific Research

Metal thread analysis provided by Mark Wypyski
Research Scientist, Department of Scientific Research

Chadar, ca. 1700. India, Mughal period (1526–1858). The Metropolitan Museum of Art, New York, Samuel D. Lee Fund, 1941 (41.19)

This chadar—used to cover the head or wrap the top of a woman's body—is an exceptional example of textiles of its type. It features a lustrous gold background with contrasting dark cotton-woven stripes and colorful silk tapestry–woven borders depicting animals, birds, and flowers within floral scrolls. Like the majority of Indian textiles woven in this style and technique in a combination of fine materials, it may have originated in the cities of Chanderi or Paithan.

Similar to Paithan saris, the borders of this textile were woven separately and then stitched together with the other parts. The chadar's components thus vary in technique—from simple plain weave for the main and end panels to tapestry weave with double-interlocking for the borders. The beige warp harmonizes with the gold wefts of the end panels and borders, while in the main panel, dark and sheer blue stripes contrast with denser stripes of silver- and gold-metal thread (fig. 1).

Fig. 1. Fragment depicting three contrasting units of the chadar: gold end panel, striped main panel, and tapestry-woven borders. Photograph by Julia Carlson

To create the multicolored borders, individual wefts of polychrome silk were double-interlocked with each other or with the metal-thread weft used in the background (fig. 2a, b). The effect is a clear, straight join on the front and a chainlike join on the back (fig. 3a, b).

Fig. 2a, b. Detail of border, front (left) and back (right). Photographs by Julia Carlson

Fig. 3a, b. Detail of tapestry double-interlocking, front (left) and back (right). Photographs by Julia Carlson

Traditionally, very fine cotton and silk were used in all types of Indian textiles. In this example, cotton is employed for the warp and for some wefts in the main panel (fig. 4). Metal thread, or zari, was used to embellish more luxurious textiles. This chadar employs two types of metal thread, composed of a strip of metal wound in a Z direction around a silk core. To ensure that the exposed area of the silk core matched the wrapping strip, and to enhance the color of the metal thread, orange silk was used in combination with gilt-silver strips, while white silk was used in combination with silver.

Fig. 4. Detail showing cotton warps and silk-and-metal wefts. Photograph by Julia Carlson

High Performance Liquid Chromatography (HPLC) confirmed that cochineal dye, made from crushed cochineal beetles (Dactylopius coccus), was used to create the silk's red and pink colors, while a mixture of safflower (Carthamus tinctorius) extracts and an unknown yellow dye (probably flavonoid dye) was used to make orange. Safflower and turmeric (Curcuma longa) were used to dye the orange core of the gold-metal thread.

Additional Reading

Barnes, R., S.Cohen and R. Crill, Trade, Temple and Court. Indian Textiles from the Tapi Collection. India Book House PVT LTD, 2002.

Dye III, Joseph M., The Arts of India: Virginia Museum of Fine Arts. Virginia Museum of Fine Arts/Philip Wilson Publishers, 2001.

Janina Poskrobko
Conservator, Department of Textile Conservation

Dye analysis provided by Nobuko Shibayama
Associate Research Scientist, Department of Scientific Research

Metal thread analysis provided by Mark Wypyski
Research Scientist, Department of Scientific Research

Carpet, late 16th–early 17th century. Iran, probably Kashan, Safavid period (1501–1722). The Metropolitan Museum of Art, New York, Rogers Fund, 1943 (43.84)

In the late sixteenth and early seventeenth century, carpets of silk and metal thread—produced in Persia by royal manufacturers and known in Europe as "Polonaise" carpets—adorned the households of the elite and often were presented as diplomatic gifts. This luxurious carpet, woven probably in Kashan, is among the few dozen surviving tapestry-woven examples from the Safavid period. They are known from the written account of a Polish-Armenian merchant, Sefer Muratowicz, sent by the Polish King Sigismund III in 1601 to Persia with instructions to order and supervise the weaving of carpets executed in silk and gold-and-silver thread. The fine silk employed for the warp and weft makes for a colorful design that emulates the style and subject matter of contemporary pile carpets, with motifs including cranes in the central medallion (fig. 1) and animals in combat in the flattened cartouches. Neighboring color fields are interwoven in a dovetailing manner—in which the wefts from both areas are wrapped and turned around the same warp alternating at regular intervals—resulting in diffuse outlines (fig. 2). Combined with the patterning of tiger's stripes and dots, these soft edges impart a sense of depth to the two-dimensional design.

Left: Fig. 1. Detail of cranes in the central medallion; Right: Fig. 2. Detail of dovetailing technique showing connecting areas of color. Photographs by Julia Carlson

Natural dyes, including red, pink, and orange hues made from safflower petals (Carthamus tinctorius), madder root (Rubia tinctorum), and cochineal beetles (Dactylopius coccus) were used to dye the silk yarn with gradations of color, even within individual sections (fig. 3).

Fig. 3. The gradation of color, called abrash, within the pink-colored ground of the cartouche is caused by differences in dye batch or a variance in the fading of the natural dyes employed (safflower petals, Carthamus tinctorius). Photograph by Julia Carlson

This carpet's beauty is enhanced by silver and gilt-silver metal thread, the product of a time-consuming, costly process in which metal wire is hammered into flat strips and wound around a silk core (fig.4).

Fig. 4. Detail of tarnished gilt-silver metal thread. Photograph by Julia Carlson

Additional Reading

Cardon, Dominique. Natural Dyes: Sources, Tradition, Technology and Science. Archetype Publications, 2007.

Dimand, Maurice S. Oriental Rugs in the Metropolitan Museum of Art. The Metropolitan Museum of Art, 1973.

A Survey of Persian Art from the Prehistoric Times to Present, Ed. Arthur, U. Pope and Phyllis Ackerman, Vol. VI. Carpets, Metalwork and Minor Arts. Meiji-Shobo/Oxford University Press, 1964–5.

Relacya Sefera Muratowicza obywatela warszawskiego od Zygmunta III krola polskiego do sprawowania rzeczy wyslanego do Persji w roku 1602. (Account of Sefer Muratowicz, Warsaw citizen, sent by Polish king Sigismund III to Persia in 1602). Warsaw, 1777.

Janina Poskrobko
Conservator, Department of Textile Conservation

Fig. 1a, b. Overall view of the front (left) and back (right) of a Child's Coat, late 19th century. India, Punjab, Amritsar or Kashmir. The Metropolitan Museum of Art, New York, The Alice and Nasli Heeramaneck Collection, Gift of Alice Heeramaneck, 1983 (1983.494.10)

Woolen textiles woven in the same technique and with similar large floral patterns as this child's coat, or sherwani, were produced by Kashmiri weavers who were active in the second half of the nineteenth century in Punjab, though it is possible that the fabric of this particular coat might have been woven earlier. The coat is tailored in the style of European men's fashion of the second half of the nineteenth century.

This delicate textile with its fine and complex pattern required a high degree of weaving expertise (fig. 1a, b). Like Kashmiri shawls, it is tapestry woven, executed in 2/2 twill weave instead of the more common plain weave. In this twill pattern, the weaver passes the weft over and under a pair of warps, creating small floats. With each successive passage of the weft threads, a diagonal alignment of the floats is formed, covering the entire surface of the textile. The area where two different colors of weft threads connect to form a double-interlocking join creates a blurry zigzag effect on the front of the textile (fig. 2a, b). The wefts of this textile are composed of thin threads made from extremely fine wool. The morphology of the fibers indicates the probable use of pashmina wool.

Fig. 2a, b. Double-interlocking technique, detail of front (left) and back (right). Photographs by Florica Zaharia

Additional Reading

Ames, Frank. The Kashmir Shawl and its Indo-French influence. Antique Collector's Club, 1986.

Barnes, R., S. Cohen and R. Crill. Trade, Temple and Court. Indian Textiles from the Tapi Collection. India Book House PVT LTD, 2002.

Dye III, Joseph M. The Arts of India: Virginia Museum of Fine Arts. Virginia Museum of Fine Arts/Philip Wilson Publishers, 2001.

Welch, Stuart Cary. India. Art and Culture 1300–1900. The Metropolitan Museum of Art/Holt, Rinehart and Winston, 1985.

Janina Poskrobko
Conservator, Department of Textile Conservation

Dye analysis provided by Nobuko Shibayama
Associate Research Scientist, Department of Scientific Research

Metal thread analysis provided by Mark Wypyski
Research Scientist, Department of Scientific Research

Fig. 1a, b. Overall view of the front (top) and back (bottom) of the Textile Fragment from the Dalmatic of San Valerius, 13th century. Spain, Nasrid period (1232–1492). The Metropolitan Museum of Art, New York, Fletcher Fund, 1946 (46.156.10)

This textile originally was applied to a dalmatic, a wide-sleeved liturgical garment, of Andalusian manufacture. It belonged to a collection of such garments attributed to the cult of Saint Valerius, the bishop of Saragossa from 290 to 315. The dalmatic was worn on the occasion of the saint's feast day. This fragment has an Arabic inscription in Naskhi script across the bottom: "Good luck and glory and exaltedness and Magnificence" (fig. 1a, b).

Woven in a slit- tapestry technique with eccentric wefts, this tapestry was made from very fine silk and the highest quality metal thread. In this technique, the weft threads are discontinuous—woven back and forth in the pattern field instead of across the full width of the textile—and a characteristic slit is created in areas where the two color fields meet. Non-horizontal wefts were used to weave curved areas and to outline them in striking colors, emphasizing the pattern (fig. 2a, b). The high warp and weft density (22 warp threads per centimeter and up to 60 weft threads per centimeter, respectively) allowed the weaver to create a fine, detailed pattern, reminiscent of miniature painting (fig. 3).

Fig. 2a, b. Detail of the weave structure; front (left) and back (right). Photographs by Janina Poskrobko

Fig. 3. Warp count per 1cm. Photograph by Janina Poskrobko

The gold metal thread used for the background is composed of a thin gilded membrane or leather strip (with an alloy measured at approximately 22-karat gold) wound around a core of silk yarn (fig. 4). Its use makes the textile rich and lustrous, perfectly suiting the Islamic aesthetic that admired luxury and refined beauty.

Fig. 4. Detail of metal thread (x100). Photograph by Janina Poskrobko

A comparison of the front and back of the textile reveals a striking difference in the intensity of red dyes (fig. 5a, b). This is a result of long-term exposure to light, especially ultraviolet radiation, which causes deterioration of natural dyes, some being more sensitive than others. In this example, the red dye obtained from a mixture of madder root (Rubia tinctorum) and a soluble red wood such as sappanwood (Caesalpinia sappan) has retained its original color on both front and back of the textile. However, the orange-red dye derived from safflower petals (Carthamus tinctorius) has faded to a light pink on the exposed front of the textile but is still vibrant on the back, where it has been well protected from the light.

Fig. 5a. View of the front illustrating the condition of two red dyes: the pink-colored weft area shows the faded safflower dye, while the thin horizontal stripe shows vibrant mixture of madder red and a soluble red wood such as sappanwood; Fig. 5b. View of the back, where red dyes derived from both sources have retained their original color. Photographs by Janina Poskrobko

Additional Reading

Von Folsach, Kjeld and Anne-Marie Keblow Bernsted. Woven Treasures-Textiles from the World of Islam. The David Collection, 1993.

May, Florence Lewis. Silk Textiles of Spain. Eight to Fifteenth century. The Hispanic Society of America 1957.

Tejidos Hispanomusulmanes, Instituto del Patrimonio Histórico Español, 2005.