Ashmolean Eastern Art Online, Yousef Jameel Centre for Islamic and Asian Art

Textile production and the use of cotton in India go back at least 4, 000 years, and may even have been established as early as the fourth millennium BCE [1]. Spindle weights were found at the Indus Valley sites of the Harappan civilization, and a cotton fibre fragment from Mohenjo-Daro, carbon-dated to 1760 BCE ±115, has been reported to show evidence of madder dyeing (Marshall 1931: 32-3; Lal 1962: 213-14). This implies the knowledge of mordant manipulation, a sophisticated technology which is the basis for dyeing with madder and other sources of red. Indian textiles may have been imported to Egypt as early as the second millennium BCE, and they were popular in the West in Roman times. Authors of Mediterranean antiquity already speculated and commented on the nature of the fibres and colour-fast dyes that produced exceptionally fine cloths [2]. Nevertheless, the early history of Indian textiles presents great difficulties, as very few fabrics have actually survived.

Reconstructing the history of Indian textiles

For a history of Indian textiles from the seventeenth century onwards, a sizeable proportion of the actual cloth is still available, both in India and abroad. For that reason, scholarship covering the later period has made great advances. Regarding all levels of production, from day-to-day wear to courtly fashion, it is often possible to match the written documentation and description with the actual textile. Representative of research done early in this century are Hadaway (1917) and Baker (1921), both studies that focused on the Indian production of painted and resist- and mordant-printed textiles. Thanks to the meticulous work of John Irwin, formerly Keeper in the Indian Department at the Victoria and Albert Museum, it was eventually possible to refine considerably our knowledge of the painted traditions, the history of chintzes in particular, and to distinguish between different production centres (Irwin 1959; Irwin and Brett 1970).

Publications of the contents of princely wardrobes and storerooms, some going back to Rajput or Mughal courts, have made it possible to gain an overview of aristocratic taste and quality of execution [3].Hence publications on Indian textiles from the time of the Mughal conquest onwards are voluminous, and they can be approached with the criteria of methodology that are applied to manuscript or easel painting. A famous early Indian source, for example, the Ain-iAkbari of Abu’l Fazl, written in 1595 as a compendium on the Mughal ruler Akbar’s government, contains information on both textiles and state revenues, as well as salaries; it has therefore even been possible to give an impression of the relative value of certain textiles in their own time (Smart 1990).

Our understanding of the distribution of trade textiles for the later period has also considerably increased, again largely due to Irwin’s research on the textile records in the India Office archives (Irwin and Schwartz 1966). His work established the production centres and trade routes from the seventeenth century onwards, and he gives us extensive lists of the names and types of fabrics used in the exchange, as well as their destination in the network of maritime trade. The production and international distribution of one specific type of luxury textile, the silk double-ikat patola cloths of Gujarat, has been covered in detail by Alfred Bühler, in particular in his and Eberhard Fischer’s monumental work of 1979 (Bühler 1959; Bühler and Fischer 1979). Bühler’s earlier major study Ikat Batik Plangi, which deals with resist-dyed textiles in general, also has important contributions regarding the use and distribution of certain textile techniques in India (Bühler 1972). Mattiebelle Gittinger’s survey on Indian cotton textile production, both for Indian use and export, has combined samples of the textiles that have survived with references in historical documents; her publication has helped to establish the outlines of a coherent picture for the Indian textile trade in the time following the arrival of Europeans (Gittinger 1982).

However, from the time prior to European involvement in the Indian Ocean trade and the trade with Asia in general, very few examples of Indian textiles have survived. Sometimes the Central Asian finds made early in this century are linked with an Indian origin, or at least with Indian techniques (Nabholz-Kartaschoff 1986: 179). However, as certain as the transmission of textiles was, either concrete evidence in the form of surviving cloth is largely lacking, or an Indian provenance is disputable. Bühler has interpreted one textile fragment found by Stein at Endere, Xinjiang (Khotan Oasis), as being Indian (Bühler 1972: i. 164). It is a blue cotton bag with resist designs, dated by Stein to the eighth century (Stein 1907: 430, 442, pl. LXXVI: E.029). Stein’s comments may also imply an Indian origin; he considered it to be tie-dyed and wrote that ‘the peculiar technique of “knot-dying”, still largely practised in North-western India, is illustrated by the piece of blue cotton cloth’ (1907: 430). Bühler correctly noticed that the technique of decoration is actually a batik resist method, rather than a tie-dye. He compared it to the representation of a head scarf in one of the Ajanta cave paintings of the fifth/sixth centuries (Bühler 1972: iii, figs. 249, 250).

It is also possible that Central Asian paintings with representations of dress may sometimes show Indian textiles, but the designs are not usually specific enough for a definite identification. Among possible ‘Indian’ motifs are small, white rosettes made up from dots and placed on a dark blue ground [4].The rosettes are indeed identical to Indian resist-printed fragments found in Egypt, but the similarities may be entirely accidental. The paintings of Dunhuang show quite specific representations of textile pattern, both in dress and in the painted ceilings which repre- sent canopies, but there is no direct link to Indian cloth [5]. On the Indian side, the paintings of the tenth/eleventh-century monastery at Ta pho, Spiti Valley, Himachal Pradesh, also reveal detailed dress designs; some of these are indeed very close to the printed textiles, although of course we have no indication as to the method of patterning for the prototypes of these paintings [6]. Actual textile fragments found by Stein at Lou-lan, Khara-Khoto, and other Turfan sites, also have rosette designs that are identical to patterns found on the Indo-Egyptian fragments, but the Central Asian textiles are made of silk (Stein 1928: iii, pl.LXXXVI). An Indian origin has to remain a purely hypothetical possibility.

Gillian Vogelsang-Eastwood has raised the question whether another important Chinese Central Asian textile find could possibly be Indian (Vogelsang-Eastwood 1990: 12 and pl. 2). It is a cotton fragment with a resist-defined design against a blue ground; it was first published in 1960 in the Chinese journal Wen Wu (Xinjiang Museum 1960: 6: 5) [7]. Hsio-Yen Shih first described it in English and also republished (from Wen Wu) the identification as an Eastern Han (25 to 220 CE) tomb find, from a double burial at Niya in northern Minfeng County, Xinjiang province. The site was abandoned in the late third century (Shih 1977: 313) [8]. Since Shih’s publication, the textile has been repeatedly published in survey books on Central Asia and the Silk Road [9]. Although the colour and weave density, as they appear from the illustration, are similar to fragments that are definitely Indian, I have doubts about an Indian origin for this particular fabric. The textile is not specifically Indian in iconography, as Vogelsang-Eastwood considers possible, but in my opinion is in part clearly Hellenistic in style. It may show a bodhisattva in a small corner square, but the manner of representation owes much to images of Gaia, the Roman Earth deity, with her cornucopia. The textile also displays a dragon tail with characteristically Chinese flames, and a fragmentary beginning of what probably was a large central field. This has a raised foot and the tassle end of a tail, as well as some curved lines that could be part of an animal’s paw [10]. In other words, it seems to me that the iconography combines Hellenistic with East Asian images, very much as was characteristic of Central Asia. The East Asian element, as exemplified by the dragon tail, certainly distinguishes the textile from the Indo-Hellenistic style of Gandhara. While it is probably true that the textile was not produced close to the burial location at Niya, a western Central Asian origin could be a possibility. Bühler considered it likely that Central Asia, especially Turkestan, early on developed its own industry of resist-dyed textiles (Bühler 1972: i. 203-5). Archaeological evidence published in 1973 has produced both cotton fabrics and cotton seeds from the Turfan basin and Khotan, of a late Han date (Sha Pi-Thi 1973).

Pfister reported three mordant-dyed cotton fragments found at Palmyra (Pfister 1937: 16, nos. 71, 72, 73, also 19-22, pl. IVd). These may or may not be of Indian origin; Pfister believed them to be so, but is also cautious about their date. Palmyra was abandoned as a settlement in 273 CE, but the cotton pieces in question may be much later, as they do not come from a clearly dated context, but apparently were surface finds. Certain textile finds from Coptic or early Islamic Egypt have also occasionally been interpreted as being Indian. However, the early reports need to be checked carefully before they are accepted as fact, as a recently carried out dye analysis has shown. One small skein of red cotton threads found at Karanis, a site in Lower Egypt excavated by the University of Michigan, was suggested to be Indian (Wilson 1933: 50). The date for the find was given as fourth/fifth century, with a cut-off date of 460 CE [11]. When the red dye was analysed in 1993, it proved to be Congo Red, a synthetic dye invented in 1885 [12]. It is unclear how the skein entered the collection of excavated items [13].

Probably Indian are two textiles illustrated by Forrer (1894: pl. I), which came from Akhmim. It is very doubtful, though, that his sixth/seventh-century date is accurate; from the material available now, it seems certain that the textiles belong to a later period. They are similar to two textiles in the Kelsey Museum at the University of Michigan, Ann Arbor (Barnes 1993: 78-9, cat. nos. 46, 47). Forrer also first published a child’s tunic from Akhmim which was resist-dyed (1898: 9-10, Pl. I). The tunic is now in the Victoria and Albert Museum (acc. no. 1522-1899) and has been discussed repeatedly in subsequent publications, of which Bühler’s summary is probably the most useful (1972:i. 161-2). He believed the tunic to be coarse cotton; this, however, has now been proven to be incorrect. A fibre analysis made at the Victoria and Albert Museum has shown it to be definitely woven from flax, and for that reason it no longer needs to enter our discussion of Indian cotton trade textiles.

Certainly Indian are two textiles illustrated by Wulff and Volbach (1926: 134, pls. 129, 130). They describe the textiles as linen cloths and give them an Egyptian provenance, Coptic or early Islamic with an eighth/ninth-century date. Direct parallels to the fragments in the Newberry Collection [EA1990.893, EA1990.894, EA1990.895, and EA1990.902] make it clear that the Berlin cloths must be cotton, are likely to be Indian, and certainly are of a much later date, probably late nineteenth century. For a similar textile in the Victoria and Albert Museum, Howell Smith already suggested an Indian provenance (1924: fig. 11).

Lamm thought that a textile fragment in the Musée Historique des Tissus, Lyon, might be Indian: it is a resist-dyed textile with floral ornaments, found at Antinoë and supposedly of sixth- to eighth-century date (Lamm 1937a: 133, fig. 62). Bidder also considers an Indian origin, but states that the textile must be of a later date, as he sees its design and technical detail closely linked to the Indo-Egyptian textiles (Bühler 1972: i. 162). I have not seen the original in Lyon, but its photograph makes me disagree and question the Indian origin. In my opinion the design is completely unlike the Indian textiles, apart from the obvious use of floral motifs [14].

To sum up the evidence from currently known material, it is clear from historical writing that Indian textiles were traded to Egypt at an early date, certainly during the Roman period. However, the actual survival of textiles is not as firmly proven as is sometimes suggested. There is no doubt, on the other hand, that the printed textiles represent a coherent group, and that they at least partly predate the European documents for the Indian trade. The results of C-14 dating on some of our fragments have confirmed this. During the first two or three decades of the century, numerous textile fragments surfaced for sale, in particular in Cairo, where they were generally associated with Fustat, the old city centre to the south of Cairo (Map 2). The fragments found in Egypt were eagerly picked up by private collectors or by museums with an interest in the decorative arts.

Pfister’s publication of Les Toiles imprimées de Fostat et l’Hindoustan

Initially the textiles were often grouped together with Islamic material (e.g. Kühnel 1927). A few early publications consider briefly an Indian origin (Baker 1921; Howell Smith 1924). But the first serious and detailed study was undertaken by the textile scholar R. Pfister; in an article published in 1936, and more extensively in a book of 1938, he put forward the hypothesis that a specific group of textiles found in Egypt was actually of Indian, and in particular Gujarati, origin (Pfister 1936; 1938) [15]. The corpus of material identified by him had in common that all fragments were of cotton, and all had been decorated by a resist- or mordant-printing technique. Usually, printing blocks had been used to stamp the resist or mordant on to the cloth.

Pfister had selected those textiles that he considered to be early, probably made before the seventeenth century. He divided all material into groups according to design motifs, for which he then tried to find parallels in Indian art. For a chronological context, he compared specific patterns with dated architectural decorations from Gujarat. Pfister also separated the designs into ‘Hindu’ and ‘Islamic’, and he assumed that the former meant an earlier date, while the latter supposedly depended on the further consolidation of Islam in Gujarat in the fifteenth century and showed strong Central Asian and Persian influences (Pfister 1938: 90). This particular division is by no means as clear-cut as he assumed, as Irwin has already pointed out:

Not only did pre-Islamic traditions persist tenaciously throughout the Islamic period in Gujarat, but there was in fact an underlying unity of regional culture which to some extent transcended religious iconoclasm, and there is no field in which this is more obvious than in the local conventions of decorative art. (Irwin and Hall 1971: 4)

Based on his comparative method, Pfister arrived at dates that ranged from the twelfth to the sixteenth century, with an increasing Islamic influence from the fourteenth century onwards. Following Pfister’s article (1936), but prior to the publication of his book, Carl Lamm also discussed the textiles in his study on medieval cotton production in Egypt (Lamm 1937a). He agreed with Pfister’s identification, but suggested a wider provenance of manufacture. Although Lamm’s investigation is brief and preliminary, and not entirely satisfactory in many respects, the question of provenance raised by him is a valid one.

Pfister’s interpretation was generally accepted, and the Indo-Egyptian textiles became known as the earliest major group of fabrics surviving from India. They were supposedly traded from Gujarat to Egypt, following the sea route past the Arabian peninsula, and then either transported overland to the Nile valley, or carried on ship north through the Red Sea.

Pfister’s argument was based on a good knowledge of the historical and art historical material available at the time. One problem remained, though: he had no firm proof for his identification of dates, other than a convincing visual link in different media. As the ornamentation of the textiles is primarily decorative rather than pictorial, iconographic peculiarities associated with any decade or even century are difficult to find or are lacking entirely. The problems this creates are quite different from those posed by an iconographic representation, as found in many of the painted Indian textiles. While the narrative style of certain of the textiles painted in south-east India often allows for very close dating, this is not the case for the designs of the Indo-Egyptian fabrics [16]. To put it bluntly, with Pfister’s stylistic analysis alone, there was no conclusive evidence that the textiles he discussed were actually of medieval date. Yet Pfister’s provisional dates have often been treated as though they were established beyond doubt.

A similar fate befell his discussion of provenance. While he had referred to the Indo-Egyptian textiles as coming from Fustat, the material became known as the Fustat textiles. Of course the two statements do not mean the same. The two questions of date and provenance must certainly be addressed again.

The problem of ‘Fustat textiles’

It has become common practice to refer to the Indo-Egyptian resist- and mordant-dyed fragments as ‘Fustat textiles’, although it has generally been acknowledged by the more careful schol- ars that Pfister’s tentative dates have to be treated with caution, to be expressed by a question-mark (Irwin and Hall 1971; Gittinger 1982; Varadarajan 1983; Nabholz-Kartaschoff 1986). Of the latter publications, Irwin and Hall as well as Gittinger reviewed the Indo-Egyptian textiles in some detail. In recent years critical comments have been raised, doubting also a clearly defined provenance (Bérinstain 1989; Barnes 1990 and 1993; Bilgrami 1990; Vogelsang-Eastwood 1990). The textiles were supposedly archaeological, yet no actual site report or identification was available from the excavations carried out early in this century.

The history of archaeological excavations at Fustat is more complex than is generally known, and therefore it may be useful to give a brief outline of both the history of the site and its excava-tions. A detailed account of its foundation and early structure as an urban centre is presented by Kubiak (1987). Fustat was founded as the new Egyptian capital in 641 CE, just after the final defeat of the Byzantine troops at Alexandria by the Arab invaders. A strong fort and the town of Babylon were at the site when the Islamic invaders arrived in Egypt, and the fortification proved to be the first real challenge to the Arabs. A long siege of more than seven months occurred, until finally the Byzantine settlement was taken. As it had been the first significant victory for the Arab army on Egyptian soil, its commander 'Amr ibn al-' As decided to found a mosque on the site after the return from the conquest of Alexandria. The Mosque of 'Amr is in use to this day, although little remains of the original building. The location of the long encampment became the new capital. It became known as al fustat, ‘the camp’; according to A.J. Butler, who studied the various names given to the site as well as their etymology, this is not an Arabic word, but has its origin in the Roman fossatum, meaning ‘camp’ at the time (Butler 1902: 340; 1914). It also continued to be referred to by its former name of Babylon, although this was given an Arabic interpretation (Bab al-Yun). The site of the old fort of Babylon is near the present-day Coptic Museum, and the small alleys and narrow streets near it are still the geographic and religious focus of the modern Coptic community in Cairo. In the space of a few hundred square metres, there are about a dozen churches. Fustat itself is just to the north of the area.

As the established town became an important urban centre, it was often simply called misr, ‘the city’. The location had clearly been a centre for trade since Pharaonic days, as an intersection for the Nile traffic with overland routes from east and west. During Roman times, a canal linked the river with the Red Sea, which added yet another area of contact. The canal now was opened again [17]. Fustat was the Egyptian capital until 969 CE, when the Fatimid dynasty established itself and the palace and government quarters were moved north to al-Qahirah (Cairo). In reality, Cairo and Fustat remained topographically connected, and the old city of Fustat continued as an economically thriving community for the next 200 years. Cairo and Fustat had become the major entrepôt for the trade between east and west, the point of interchange between the Asian and Mediterranean economies.

However, in 1168-9 Fustat was largely destroyed by fire and briefly abandoned, to avoid capture by the advancing Crusader army that tried to conquer Cairo. Although under Saladin a certain amount of resettling occurred, the old city quarters apparently never fully recovered. By the end of the thirteenth century many parts of Fustat were largely a wasteland, with the exception of settlements along the river. Beginning in Mamluk times, parts of the formerly thriving urban site became the rubbish disposal area for the growing city to the north. Once the hub of the trade of Asia with the Mediterranean, Fustat is now a wasteland of rubble and rubbish [img]1[/img] [img]2[/img]. Squatters have built shacks on the edge of it, and they make a living by picking over the refuse and- surprisingly- by making and selling crude earthenware pots, a product for which the area has been known since Mamluk times. The kilns are lit by the rubbish around them, and huge clouds of billowing smoke from burning rubber pollute the air and turn the sky black. In between are the remains of archaeological excavations.

Despite the abandonment of Fustat as an urban centre, the Coptic and Jewish communities continued to use their houses of worship in the old quarters. It is thanks to one Jewish synagogue that we have extensive documentary evidence for the lives and activities of the Jewish community of Fustat, during the Fatimid period in particular. Many medieval synagogues had a depository for paper that had the name of God written on it, called a Genizah. Usually the accumulated material was regularly disposed of [18]. In the Fustat synagogue, however, the papers were not buried, but were all kept in a separate chamber, a room that was not opened until c. 1890, when the synagogue was renovated. By that time, the paper scraps of almost 1, 000 years had accumulated. S. D. Goitein’s name in particular is associated with research on the content of the papers (Goitein 1967, 1971, 1978, 1983, 1988).

The first archaeological excavations were at the fort of old Babylon, which was of Roman origin. Butler thought that the fort was built at the time of Trajan, c. 100 CE, but adds that it only replaced an older fortification (Butler 1902: 243-4). From Butler’s description it does not actually seem as though these first excavations were very archaeologically inclined, but were rather a clearing operation. He mentions the name of Max Hertz, who excavated the Roman gateway and cleared away mounds of rubbish (Butler 1902: 239).

More serious archaeological work in Fustat was begun by Aly Bahgat under the auspices of the Arab Museum in Cairo, later renamed the Islamic Museum (Bahgat and Gabriel 1921). Bahgat’s excavations are now viewed with a certain degree of scepticism, in particular regarding his chronology. In the publications about these excavations, no printed textiles are mentioned, Indian or otherwise. As the large site of Fustat (not necessarily excavated by Bahgat) has been a refuse disposal area for centuries, its frequently disturbed surface has made the dating of finds very difficult. It seems to me most likely that the Indian textiles have come from this general refuse disposal, rather than specifically from parts of the excavations carried out by Bahgat; the first finds may even have come specifically from the clearings of the mounds around the Roman fort. Certainly none of them can be proven to have come from medieval graves, at Fustat or elsewhere [19]. In the 1930s some graves were opened in the Fustat region, however. They produced a number of tiraz textiles. The site of Fustat, therefore, poses many questions, and even the relatively recent excavations under the direction of George Scanlon did not provide conclusive answers for our material. His work started in 1964 and continued to 1980 (Scanlon 1986; Scanlon and Kubiak 1989) [20]. In the introduction to the final report of the last season at Fustat-C, there is an apt description of why the excavations were coming to an end: the rubbish disposal of the city of Cairo was finally pushing out the archaeologists.

Work does continue in other parts of the area, however. A French team under Pierre-Roland Gayraud has been excavating near the Islamic cemetery in the southern part of Fustat during the last decade, so far without any published records on textile finds, although many fabric fragments have actually been found at the site [21]. During the last excavation season under Scanlon’s auspices, however, in 1980, and referred to as Fustat-C, textiles were actually recorded: more than 3,000 fragments were found, but of the textiles analysed by Louise Mackie in Toronto, only one is likely to be Indian. Based on coin finds associated with its location in the excavation, it was given an eleventh-century date (Mackie 1989: 88-9, frontispiece 2). This relatively early date for what might be an Indian fragment is of great interest in light of the C-14 analysis of fifteen of the Newberry fragments, of which two were also of eleventh-century date. The following chapter will discuss the possible connections.

To sum up the history of archaeological research, there are three excavation projects that have been recorded in detail, under the direction respectively of Bahgat, Scanlon, and Gayraud. In addition there is the 1930s opening of graves in Fustat, as well as several excavations continued by the Islamic Museum and the Egyptian Department of Antiquities, with no published results. Finally, there has been a constant sifting and rummaging of the disposal area by people collecting sibakh, organic material that has become nitrogenous and can be used as fertilizer. Of course the same people may on occasion turn into ‘treasure hunters’ seeking objects that could bring some money if offered to Cairo’s antique dealers. At the moment we may assume that the Indian textiles came on to the market by the latter way.

Fustat itself, therefore, can only provide limited answers to the questions of provenance and date. We have no proof that all of the Indian fragments have actually come from the site; they may have arrived on the art market from a different source. As, apart from Fustat-C, no recording of the archaeological context of the finds is currently available, the dates one can arrive at on stylistic grounds are also in no way definite. The textile finds from Fustat-C tell us one thing, though: much of the textile material associated with Fustat itself, as a former centre of Mediterranean as well as Indian Ocean trade, is actually not Indian, but was either produced as part of the very important Egyptian textile industry, or traded from elsewhere in the Mediterranean or Middle Eastern realm.

Quseir al-Qadim

In 1978 Donald Whitcomb and Janet Johnson from the Oriental Institute in Chicago started an archaeological exploration at the Red Sea port of Quseir al-Qadim. Excavations continued until 1982, and the preliminary reports have appeared in recent years (Whitcomb and Johnson 1979; 1982). The site is located about eight kilometres north of the small modern port of Quseir. It had been in use as a trading port in Roman times and again from the thirteenth to the fifteenth century, at the time of the Ayyubid and Mamluk reign in Egypt. By the early fifteenth century, the harbour was beginning to silt up, and the location lost its importance as a trading port. The settlement apparently shifted to the modern location of Quseir after that time.

The two periods of occupation at the archaeological site coincide with times of very active direct trade between Egypt and the east. The Roman finds are dated to the first and second centuries CE, when the port was called by its Greek name Leucos Limen or, in Latin, Albus Portus, both meaning ‘white port’ (Murray 1925: 141) [22]. During neither the Roman nor the Ayyubid/Mamluk period was Quseir the major Red Sea port, but it was the one with the shortest overland route between the sea and the Nile valley. Travelling through the Wadi Hammamat, there were approximately twelve Roman watering stations along the 180 km. road to the Nile valley town of Coptos (Quft, just north of Qus). During the Islamic period, Quseir was the Red Sea port for Qus, since Fatimid times the capital of Upper Egypt.

Because of the difficulty for large ships to navigate in the Red Sea, the southern port of Aydhab, near the modern settlement of Suakin south of Port Sudan, was considerably more important for the distribution of international trade during the Mamluk period. But Quseir did receive its share of the eastern trade. Qalqashandi wrote in the fourteenth century:

al-Quseir is on the northern side of Aydhab and some of the ships frequent it; it is near to Qus and Aydhab is far from Qus. The merchandise is carried from Quseir to Qus, then from Qus to the warehouse of al-Karim in Fustat. (cited in Whitcomb and Johnson 1979: 4)

The port is in a very dry, desert-like location, without any agricultural hinterland. For this reason, it was completely dependent on its life-line to Coptos or Qus. The only food available locally was fish from the sea, and there are abundant remains of fishing implements among the finds.

The archaeological finds show that Quseir in Mamluk times received Chinese porcelain and celadon ware, as well as textiles. For the Roman period, glass and cloth were present, at the time the major export articles from Egypt to the east. The glass trade for both the Roman and Islamic periods has been analysed by Carol Meyer (1992). Of great interest from the Roman period are pottery finds that clearly link Quseir with India. Most remarkable is a storage jar discovered in 1980, with a Brahmi inscription identical to the script found at Arikamedu, in southern India, commonly used to write Tamil, and dating to the first and second centuries CE, in complete agreement with the Quseir context (Johnson, in Whitcomb and Johnson 1982: 263-4) [23]. Arikamedu was an entrepôt and port site near Pondicherry on the south-eastern coast of India, active as a trading station between east and west in the first century CE. The inscription at Quseir was similar to a graffito found in the same area in 1978. A comparison of the pottery finds from Quseir, in particular from this specific location of the inscription finds, with the pottery present at Arikamedu, revealed similar first- and second-century Roman sherds. Furthermore, Quseir al-Qadim had pottery examples that were identical to what had been called native Indian ware at Arikamedu (Whitcomb and Johnson 1982: 7). Not only does this confirm the contact with the south-east coast of India, but the graffiti also suggest that Indians may have been present at Quseir.

Among the material found from the second period of habitation are 68 cotton fragments that bear direct relation to the textiles Pfister identified as coming from Gujarat, both in design and technical details (Eastwood 1982; Vogelsang-Eastwood 1990). They are resist- and mordant- dyed textiles, mostly block-printed. The dyes identified are indigotin and/or a red colorant, provided either by madder (Rubia tinctorum L.) or a variety of morinda roots (Morinda citrifolia L. or Morinda tinctoria Roxb.). The same dyes were analysed for the textiles in the Kelsey Museum, University of Michigan (Barnes 1993: 92-7) [24]. Madder, morinda, and indigotin were also identified in the Newberry fragments. The Quseir find means that for the first time we now have sufficient evidence for the presence of textiles of this sort from a site that was excavated under datable conditions.

Certain of the fragments found at Quseir have exact parallels in the various collections of Indo- Egyptian textiles, including, of course, the Newberry textiles [img]3[/img] [img]4[/img] [25]. The report of the find makes the Quseir al-Qadim excavations the single most important discovery for the study of our textiles. This is especially the case as the published reports give the details of location, as well as a reliable account of the specific and general context. Archaeological work at Gebel Adda, with a thirteenth-century date, and Qasr Ibrim, without dates so far, also has produced comparable material. Some of the Gebel Adda finds are briefly discussed by Vogelsang-Eastwood (1993b). Most recently printed textiles have been reported from Coral Island in the Gulf of Aqaba south of Eilat. The site is also identified with the Ayyubid and Mamluk period, with a twelfth- to fourteenth-century date (A. Baginski, personal communication) [26].

One fragment from Qasr Ibrim in Upper Egypt is virtually identical to the border of tabs on a large textile in the Newberry Collection [EA1990.1073]. The chronology for Qasr Ibrim is generally very difficult to establish, as much of the material comes from surface finds. But the Qasr Ibrim textile clearly shows that identical Indian textiles were not only traded to the north from the Red Sea ports, but also south into Nubia. To sum up, the archaeological evidence shows that Indian textiles, in particular block-printed and resist-dyed, were commonly traded to Egypt by the thirteenth and fourteenth centuries, and that they were not only in use in metropolitan centres, such as Fustat, but had a wide distribution.

It is against this background of recent discoveries and research results that the Indo-Egyptian fragments in the Newberry Collection can be reappraised. Our own contribution to further research lies in particular in the addition of C-14 results to the conventionally dated archaeological records. Following Pfister’s ground-breaking publication, these discoveries offer the first substantial breakthrough in our understanding of the early trade with Indian textiles. From the evidence of their distribution, which is still scant, it is obvious that the Indian textiles collected in Egypt should be newly defined, and that the term ‘Fustat’ should be abandoned as their designation.