Tapestries produced after the Spanish Conquest of Peru in 1532 form a hybrid group of artworks that spans the cultural and technical horizons of the age of exploration and discovery. The Museum is fortunate to have several in its collection, including one masterwork featuring a quintessential mixture of Andean and Spanish aesthetics and manufacturing techniques (Figure 1). As this tapestry will be featured in the upcoming exhibition "The Colonial Andes: Tapestries and Silverwork, 1530–1830" (September 27–December 12, 2004), it is currently undergoing study and conservation.

Woven of yarn made from indigenous fibers and colored with locally produced dyes, this highly detailed tapestry was influenced stylistically by European grotesquerie of the late sixteenth and early seventeenth centuries, which was artfully adapted by Andean weavers. Figural motifs derived from European sources, including the Old and New Testaments and Classical mythology, are intermixed freely with local imagery, such as depictions of Andean nobility from pre-Inca times (Figure 2). One of the rare colonial textiles that bear writing of any kind, the tapestry includes the enigmatic inscription "Moussom Nessept." Also unexplained is the meaning of the central roundel, with its amorphous forms containing winged faces, that may be an adaptation of a Chinese cloud motif derived from silks arriving from Asia via the Manila Galleons (see Figure 10).

Cumbicamayos and acclacuna, the Inca weavers from the highlands of Bolivia and southern Peru, were organized in cloistered workshops, where they refined their skills to unprecedented heights. The technique of tapestry weaving was reserved for the production of cumbi, the finest cloth used only for royal garments, but this practice changed after the arrival of Spanish administrators and clergy, who commissioned large tapestry-woven wall hangings that were customary in Europe. Even as the size and format of their woven products changed from garments to tapestries, Andean weavers continued to use their traditional techniques and materials (Figure 3), only occasionally adding to their work the silk and metallic threads that had been introduced by the Spanish.

The Metropolitan Museum's tapestry was woven as a single, four-selvage, double-faced unit. For the most part, painstaking single interlocking joins were used to weave together adjacent fields of different colors. The weavers used additional methods, such as dovetailing and eccentric wefts, to join separate fields, but also as a decorative effect within monochrome sections. In addition, several figurative designs have been embellished with three-dimensional accents in a weft-float pattern (Figure 4). Unlike their European counterparts, Andean weavers cleanly finished both the front and back of their tapestries, rendering textiles, as observed by a Spanish chronicler, "with two faces." The division of the tapestry into a series of vertical fields that do not correspond to the design is a puzzling feature seen on many colonial Andean tapestries (Figure 5).

While the varying levels of skill that were applied to the manufacture of the tapestry suggest multiple hands, these individual sections are too narrow to be the work of different weavers. In 1967, Adolph Cavallo, in Tapestries of Europe and Colonial Peru from the Boston Museum, hypothesized that the vertical bands are a misinterpretation by local weavers of seams in Spanish knotted pile carpets composed of narrow, loomed widths sewn together.

The warps of the tapestry are three Z-spun, S-plied cotton yarns, while the wefts are two Z-spun, S-plied yarns made of wool, presumably from alpaca and vicuña, two of the native camelids of the Andes. The finest hair from these animals is extremely soft and was prized by the Andean peoples as well as by the Spaniards, who likened it to silk. After the Conquest, sheep were imported to the New World, but their wool is coarser than that of camelids, and it seems not to have been used by Andean weavers for their finest products. Because the macroscopic and microscopic features of camelid hair vary according to species, gender, age, and the location of the fibers on the animal's body, differentiation between wools from diverse sources can be difficult, even in the laboratory. In this case, even details revealed by electron microscopy did not allow conclusive identification of the fibers. Future investigations may benefit from new analytical methods now being developed, such as amino-acid array testing.

Three-ply warp yarns were used by Inca weavers in the production of cumbi, which required very strong warps to withstand the high stress of such a fine, tightly woven weft. The tapestry's warp and weft counts—respectively 26 and 148 wefts per inch—illustrate one reason why the Spanish were so impressed by the quality of Inca weaving. In comparison, the thread counts of medieval European tapestries, such as the famous Hunt of the Unicorn series on display at the Cloisters, range from sixteen to eighteen warps and sixty to ninety-six wefts to the inch.

The colors preserved in the Museum's tapestry form a limited, yet familiar palette. Wool was used in colonial Andean textiles in all of its naturally occurring shades of white, brown, gray, and black, but fibers were also colored red, blue, and yellow with dyes derived from a variety of native plants, animals, and minerals. In order to identify the dyestuffs used on the tapestry, analysis was conducted with HPLC (see High-Performance Liquid Chromatography). Carminic acid, the main color compound in cochineal dye, was found in four red yarn samples of different hues. A scale insect living on cacti in various regions throughout the highland valleys, cochineal (Dactylopius coccus), is the source of this brilliant red dye (Figure 6). The Spanish encouraged its production, and the huge quantities exported to Spain were second in value only to the shipments of silver coming from the great mines.

Indigo was identified as the colorant in a sample of blue yarn. Before the Conquest this dye was derived only from local plant varieties (Indigofera spp.), but seventeenth-century Spanish sea traders also brought indigo to Peru from more northern colonies such as Guatemala, where indigenous indigo plants had a higher yield. The chromatogram obtained from the extract of a yellow yarn sample, reconstructed at wavelength 380 nm, presented only one minor diagnostic peak, at twenty-four minutes. This may indicate that some of the colorant degraded over time, leaving deterioration products that do not have a homogenous structure. Based on its retention time and absorption spectrum, the small peak appears to be that of rutin, a flavonoid (O-diglycoside) derived from various plant species.

Preparation of the tapestry for exhibition started in 2001 with a thorough examination and documentation of its condition. Color preservation of the dyed woolen warp threads is generally good, with only slight fading. Structurally, the tapestry is surprisingly sound, although in small areas damage was caused by insects, and repeated handling has resulted in mechanical wear at the selvages (Figure 7). The original cotton warps still provide structural support, while in general the weft threads are also well preserved. Modulation in the weaving around figures sometimes resulted in wavering patterns that give the tapestry a distinctive texture and visual vibrancy, but also adds a physical elasticity that has contributed to the work's remarkable state of preservation (Figure 8).

Nonetheless, few centuries-old works of art have escaped restoration, which in this case was carried out before the textile entered the Museum's collection and had focused on improving structural stability, for the most part by reweaving damaged areas. This fairly crude remedy did prevent further damage, but unfortunately distorted the tapestry's original structure and marred its appearance. Thick cotton threads with short fibers and a low degree of twist were used to replace the original warps, weakening the structure of the weave and producing restored areas that were overly rigid. The wool threads used to replace the original wefts also were too thick, and were colored with unstable, synthetic dyes. The resultant fading caused discontinuities within motifs, which severely affected the visual integrity of the entire work (Figure 9). Most significantly, in disregard of the fact that the tapestry was conceived as double-sided, an obverse and reverse were effectively established by finishing the rewoven sections only on one side.

In anticipation of a long treatment period, the most critical areas were addressed first, after which the work could progress as time allowed. Based on several factors, including the importance of this unique work to the collection and its overall condition, the decision was made to clean the textile using a vacuum and to replace previous restorations. Although this approach has been applied in the Department of Textile Conservation for several decades, in particular to medieval European tapestries, the extraordinary qualities of the Andean weaving created an unusual challenge. Not only was the visual effect of the tapestry much improved by eliminating conspicuous patches of earlier restorations (Figure 10), but as a result of reweaving these areas, the fabric regained much of its original elasticity as well.

Given the pictorial nature of the tapestry, it was important to assure not only the physical but also the visual connection between original and newly introduced materials. Reconstruction of the imagery could be based on evidence surviving in the areas adjacent to the damages, although in some instances similar motifs from contemporary works had to be studied. Compared with European wall hangings, on which most figures are defined with fine brown outlines, the contours of the images on this Andean work are emphasized with broad borders—approximately two millimeters in width—woven in a contrasting color and, as such, the recreation of accurate line thickness, orientation, and color were essential to the success of the treatment. Reweaving was complicated by the use of a variety of structural techniques and many different hues, particularly within the red-pink background. Because the original texture is so fine, only a single thread could be used, which created difficulties in matching original nuances in the original. To increase color variation within "monochromatic" fields, two or three needles with yarn of different hues were used in alternating fashion. The six vertical bands mentioned earlier often showed discontinuities in color where they were joined (Figure 5), and this characteristic was reproduced in the conservation treatment as well.

Among the features that indicate the work of weavers with different levels of skill are the concentration of wefts and shifting of warps in the Museum's tapestry, which appear to be greater on the right side than on the left. The presence of these different hands required that their individual styles be emulated in the reweaving, by giving special attention to the tension within reconstructed areas.

Condition permitting, tapestries are generally displayed vertically to allow museum visitors to view them from an optimal vantage point. As a rule, for mounting purposes, strap supports are applied to the back, in order to reduce stress on the warps and wefts, which are generally weakened due to fiber fatigue. In some cases a lining also is sewn onto the reverse, a measure that provides extra support while at the same time preventing the accumulation of dust. Unlike its European counterparts, the Andean tapestry hangs in the warp direction, and its excellent condition allows it to be suspended from a single strip of cotton webbing, prepared with Velcro^®, hand-stitched across the top edge. This approach provides the additional advantage that access to the reverse is preserved. For storage after the exhibition, the tapestry will be removed from its mount and rolled onto an acid-free tube that rests in an archival cradle. The cradle is housed at the Antonio Ratti Textile Center storage facility, in a climate maintained in accordance with conservation standards.

The dyes used in the Andean tapestry (see Figure 1) were analyzed with the use of high-performance liquid chromatography (HPLC) using a photodiode array (PDA) detector (Figure 11). HPLC is an analytical technique that has varied applications in industry, but since the late 1970s it has been used in museum laboratories to identify or characterize organic compounds found on works of art. Chromatography is a technique for separating mixtures of compounds, introduced in a mobile, liquid, or gaseous phase, on the basis of their physical or chemical affinity to a stationary medium. In the case of HPLC, a dissolved sample is forced under high pressure through a column containing an adsorbent packing material. As a sample is separated into its individual compounds, the moment at which each elutes—the so-called retention time—is recorded. The PDA detects the absorption spectra of the individual components in the ultraviolet to visible-light range (200–800 nm). These characteristic spectra are used in combination with the retention times to establish the composition of the mixture (Figure 12).

For the purpose of identifying colorants used in the Andean tapestry, dyes were extracted from one yellow yarn and four red yarn samples, each of a different hue, using 400 µ-1 ml of 1N hydrochloric acid/methanol (2:8) in a dry-bath incubator (80–90ºC). The extracts were evaporated, dissolved again in 10–20 µl formic acid/methanol, and aliquots of each solution were injected into the HPLC. Dye was extracted from a blue yarn with methylene chloride in formic acid. The extract was dried and the residue redissolved in 20 µl methylene chloride/formic acid (1:1), of which 5 µl was injected into the HPLC. For comparison, reference samples colored with known natural dyes and color-compound standards were also analyzed.

The analytical system consists of a 1525 µ binary HPLC pump, a 2996 PDA detector, a 1500 series column heater, and an in-line degasser, all produced by Waters, and a Rheodyne 7725i manual injector. A Waters Xterra RP18 3.5 µm 2.1 x 150 mm reverse-phase column was used with a flow rate 0.2 ml/min. The binary mobile phase consisted of (A) 2% acetic acid in a 0.1M ammonium acetate deionized water solution (v/v) and (B) 2% acetic acid in methanol (v/v). Separations were achieved by a series of linear gradients of B to A as follows: elution starting with 10% B in A, < 1 min. isocratic; 10–30% B in A, 1–6 min.; 30% B in A, 6–16 min. isocratic; 30–60% B in A, 16–21 min.; 60% B in A, 21–31 min. isocratic, 60–90% B in A, 31–36 min.; 90% B in A, 36–50 min. isocratic. The column temperature was 30ºC. The operation and data processing software was Empower Pro (2002).

Elena Phipps received her Ph.D. in Art History and Archaeology from Columbia University. She is Conservator in the Department of Textile Conservation, where she has worked since 1977, and is currently co-curator for the upcoming exhibition "The Colonial Andes: Tapestries and Silverwork, 1530–1830."

In 1988 Florica Zaharia joined the Department of Textile Conservation, where she was recently named Acting Associate Conservator in Charge. She earned a Ph.D. in Visual Arts from the University of Art Nicolae Grigorescu in Bucharest, and has contributed to many major tapestry conservation projects, including treatments of The Courtier with the Rose and The Start of the Hunt of the Unicorn.

Nobuko Shibayama is Associate Research Scientist in the newly established Science Group. She holds a Ph.D. in Applied Science for Functionality from the Kyoto Institute of Technology, where her thesis focused on the separation and identification of natural dyestuffs in archaeological fabrics using liquid chromatography/mass spectrometry. In addition, she received a Diploma from the Textile Conservation Centre in the United Kingdom.