TEXTILE REMAINS FROM A LATE TEMPLE IN EGYPTIAN NUBIA
Nettie K. Adams
In 1986 the Egypt Exploration Society1 discovered a collection of ex traordinary archaeological textiles in Egyptian Nubia (Figure 1). The textiles were excavated from a previously unknown temple found in the walled, fortress-city of Qasr Ibrim (Figure 2). Located high on a bluff overlooking the east bank of the Nile, Qasr Ibrim dominated the area before the Aswan High Dam flooded the valley. Its commanding position is gone today, but most of Qasr Ibrim still remains above the waters of Lake Nasser. Occupied continuously for more than three-thousand years, Qasr Ibrim was inhabited at least as far back as the time of Ramses the great, around 1250 B.C. Subsequent groups who established themselves were successively the Kushites, from farther south along the Nile; the Ptolemaic rulers of Egypt, and then the Romans, who garrisoned the site in 23 B.C. The Romans were supplanted by people from the city of Meroe, deep in the Sudan. This period is known as the Meroitic pe riod, and lasted at Qasr Ibrim from about 100 A.D. to about 350 A.D. Although African in origin, the Meroites shared many aspects of the general culture of the classical world. The Meroitic empire collapsed around 350 A.D., but many traits of Meroitic culture survived at Qasr Ibrim for the next two-hundred years. This period, from 350 to 550 A.D. is the time frame for the textiles found in the temple (Figure 3). Although built probably during the Egyptian twenty-fifth Dynasty, about 750 B.C., the temple remained in use for thirteen-hundred years, until Christianity came to Nubia around 550 A.D. Evidence suggests that its destruction was brief and violent. Altars were over turned and statues were smashed - actions consistent with Christian ity's commandment against idols. Pottery and stone objects, vessels of faience and glass - all were ravaged. When the destruction of the tem ple contents had run its course, the mud brick walls were pulled down, forming a protective cap over the damaged furnishings. The roof beams were removed and doubtless used elsewhere; smoke stained
ARS TEXTRINA 8 (1987), pp. 85-124 pieces of matting from the ceiling were found in with the fill of broken brick. The remnants of the temple were then covered with additional fill and the area was used as a plaza adjoining the cathedral, which was soon to be built nearby.2 This description of the vandalism and destruction wrought upon the temple objects holds for the furnishing fabrics of the temple as well. Extraordinarily fine textiles were wilfully torn into pieces and the fragments scattered throughout the several rooms of the temple. Ul timately, one-hundred and fifteen textile fragments were identified as having been part of the temple furnishings; from these were recog nized twenty separate specimens, of which eight were assembled or reassembled. Considered as a group, there are certain features which are worth noting. First, although the use these textiles served in the temple is in question, it does not seem likely that they were garments. There are no signs of cutting or shaping, and the only sewing, except for the con tainers, was for the purpose of mending. Second, all but three were made of cotton. This is interesting in view of the flax and wool in gen eral use in Egypt of the same period, but not unusual for Nubia. All yarns are spun in the "s" direction, continuing a tradition going back to the beginning of weaving in Egypt. Where plied yams occur they are always s-spun, z-plied. The weaving techniques employed are limited to plain weave, with one, and occasionally two weft yarns passing over one warp; and tapestry weave, which permits elaborate patterns to be created. Another feature common to this collection is the limited color range. Again, with one exception which will be discussed later, the only colors found are dark blue, a medium sky blue, pale blue, and the natural color of the undyed cotton. Furthermore, the colors are so ar ranged that the shades of blue are almost never in direct contact: they are separated by white or undyed threads.3 This seems to be compara ble to the law of medieval heraldry, which forbids contact between red and blue. Finally, a condition shared by the furnishing fabrics is that they still retain a slight fragrance of the incense which must have been used in the temple. This feature provided an unexpected dimension to our reconstruction of the activities and rituals of some fifteen-hundred years ago.
86 hi archaeology, the place where each object is found is of paramount importance: it is this provenience which permits the inferences and interpretations so vital to understanding what happened in ancient times. My organization of the textiles discussed below has been deter mined primarily by this principle; sub-groups are based on technical characteristics. The two major groups of textiles found within the temple are the furnishing fabrics and the textile miniatures. The fur nishing fabrics can be organized into five groups; curtains, containers, bordered cloths, tapestry-woven fragments, and decorative elements. The miniatures, found within the temple sanctuary, were not part of the temple furnishings. They will be discussed finally.
Furnishing fabrics.
Curtains Curtains, so designated because of their relatively large size and the lack of any sewing or shaping, are all of cotton plain weave, and all were dyed blue after weaving. None showed signs of wear, but all had been deliberately torn into pieces and scattered about. One dark blue specimen (Table 1,1; Figure 4) has been assembled from fifteen frag ments. It seems clear that only a small portion of the object was recov ered; just two sections of one selvedge were found, all the other bor ders are missing so that the original size of the piece is impossible to reconstruct. A second curtain (Table 1, 2) is similar to the first, but is a slightly lighter blue. It has been assembled from twenty-three fragments. A third specimen (Table 1,3; Figure 5) of pale blue, is a more delicate fabric than the first two, having finer yarn and a more open weave. It retains one selvedge and an end border of wrapped openwork (Figure 6). This decorative end border technique is commonly found on textiles from Meroitic and X-Group levels at Qasr Ibrim (Crowfoot 1984: 10- 17); it appears also at Karanog (Randall-Maclver and Woolley 1910: Vol. IV, Plate 108, Figure 5), at Gebel Adda (Crowfoot n.d: 5) and at Qustul (Thurman 1979: 40-41)4. Like the other two specimens, the
87 thirty-two fragments identified as part of this curtain are mute evi dence of wanton vandalism. Although we have no idea where in the temple this curtain originally hung, it seems possible that it may have been a parapetasma , which has been described as the "precious curtain which forms an essential part of the setting for a ritual sacrifice" (Weitzmann 1972:6). Such a cur tain appears in the upper right corner of a carved ivory plaque of an Isis cult scene, now in Dumbarton Oaks (ibid. 5-7, Plate 1). The plain weave of our curtain does not seem very special to our eyes, but to the ancients, a solid blue cloth must have been precious indeed, since most of their fabric was undyed.
Fabric containers Three objects, principally of textile construction, which were found in the temple fill, can be identified as containers. The most structurally complex is a double-chambered case (Table 1,4; Figure 7), which may originally have been fitted with glass vessels or other fragile objects. This case was constructed of two cylindrical basketry structures cov ered by plain blue cotton cloth and joined together side-by-side. Each cylinder was originally provided with a round bottom, and also with a now-missing lid which was attached to the case by a decorative plait, which served as a hinge. These attachments are still present; the hinge is anchored to the case at the back with a stout knot, and on the front, a second knot is present which would have secured a loop on the missing lid. (A lid, probably similar to the ones for this case, was found in ex cavations several meters south of the temple.) The case was decorated with an all-over couched pattern resembling netting, and with decora tive plaits similar to the hinge strip. The second container is a long, narrow bag (Table 1, 5), found in one of the temple crypts. Its top is very deteriorated from long handling and possibly from being rolled or folded down. It is made from a sin gle piece of cloth, folded and stitched along one side and across the bottom. The fabric is a complete piece - that is, it has both side and both end borders intact. Decoration was rendered in bands of medium blue and dark blue on an undyed ground.
88 The last, and most problematic fabric container is a linen-lined leather fragment (Table 1,6). One long edge appears to be finished; the other edges are ragged and incomplete. The lining was composed of three layers of plain-weave undyed linen, attached to the leather by bold decorative stitching in flax thread. Elaborate patterns created by the decorative stitching mark it as a cover for a valued and important ob ject: it might have been a bag or case to hold a sacred or precious item. The flax lining is one of only three exceptions to the use of cotton in the collection of temple textiles.
Bordered cloths The most extraordinary group of temple textiles to be examined are three cotton bordered cloths which are like tapestry-woven pictures. These three cloths vary in size5, but are structurally similar. All of them have a dark blue border which acts as a frame surrounding the pattern. This framing border is itself surrounded by an undyed ground weave with paired wefts. Instead of being square, the four corners of each framing border have a small triangular projection which protrudes into the ground weave. Two of the cloths were mended in antiquity. Within the dark blue framing border of the smallest cloth (Table 1, 7; Figure 8) is a central field, and a second field which surrounds it. The pattern is woven in such a way that the central field appears to be su perimposed over the outer field. Within the medium blue central field are seven horizontal dark blue rectangles, outlined with a narrow white line. Their arrangement is staggered, with three in one row, four in the other. Inside each rectangle are two white horizontal ankhs lying side by side. The outer field has a diamond pattern made of dark blue stepped squares on a white ground. Within each diamond shape is. a single horizontal ankh in white. Narrow lines of dark blue and white border the edges of the outer field. Along the inner edges of the dark blue framing border is a single row of alternately larger and smaller white dots. The middle-sized bordered cloth (Table 1, 8), in addition to the ground weave and framing border found also in the smaller one, has two additional borders surrounding a single field. The field is similar
89 to the one just described, except that the motif inside each dark blue rectangle is a quatrefoil suggesting a flower or Greek cross. The rect angles number thirty-eight, and are arranged horizontally in four staggered rows of alternately nine and ten each. At the outer edge of the field is a narrow dark blue border with a single row of alternately larger and smaller white dots. Between the framing border and the inner border is another border of meanders. This motif is similar to one illustrated from Karanog (Randall-Maclver and Woolley 1910: Plate 108, Figure 3). The largest and most complex of the three bordered cloths (Table 1,9; Figure 9, 10) has the double field arrangement seen in the smallest cloth. Both fields have the diamond pattern made of stepped squares, distinguished only by different use of the three colors. The outer field has the same color relationships as the outer field of the smallest cloth: dark blue on a white ground. The diamond pattern of the central field is made of sky blue stepped squares, completely edged with a narrow white line, against a dark blue ground. The motif inside each diamond shape of both fields is an eight-pointed star: white in the central field dark blue in the outer field. A small dot in the center of each star is of the opposite color. Between the two fields are narrow lines of white and dark blue; the same occurs in reversed order between the outer field and an inner border. The inner border, one centimeter wide, has a dark blue ground through the center of which runs a row of sky blue dots joined side by side. Even these have been very carefully outlined in white to separate the two shades of blue. The dark blue framing border is in four pieces, but, very fortunately, one of them still retains a corner of ground weave with part of a selvedge and a portion of end border having a few remaining stumps of wrapped openwork. This specimen is one of extraordinary technical quality. The even spinning of the yarn and the regularity of the weave make this the finest exam ple of textile craftsmanship yet recovered from the late classical period at Qasr Ibrim.
Tapestry-woven fragments Twenty-two cotton textile fragments, assembled into six specimens, comprise a group referred to as tapestry fragments because all are very fine examples of tapestry weave. The first specimen (Table 1,10)
90 in two pieces, has a repeating pattern of closely spaced ankhs The weaving is very fine and close, having thirty-six weft yarns per centimeter. The dark blue ground forms a rich backdrop for the ankhs which were originally almost white; their color has now aged to a tawny, light brown. Another specimen (Table 1,11) represented by a single fragment, also has a small, repeating tapestry-woven motif on a dark blue ground. In this case the motif is more widely spaced on the background. It is a dot with an attached hook, woven in the tiniest possible scale (0.8 by 0.3 cm), and probably represents a ladle 91 A similar tapestry-woven pattern is seen on two other, more frag mentary, specimens (Table 1, 13, 14). Neither has the openwork bor der nor the ground weave of number 12, but number 13 has the same eight-pointed star motif; number 14 has a small white quatrefoil simi lar to that of one of the bordered cloths (Table 1,8). A similar border pattern was also found at Karanog (Randall-Maclver and Woolley 1910: Vol. IV, Plate 108, 2). Fifteen fragments from three separate proveniences were assembled for the final specimen in this category (Table 1, 15; Figure 12). It is technically one of the most expert, and artistically one of the most striking of all the temple textiles. It consists of a row of ten figures termed the Meroitic sa 6, each measuring 4.5 by 8 cm, boldly rendered in white, against a dark blue ground. In the spaces formed by the loops, the color alternates between medium blue and white. Where it is preserved, a solid bar in white bounds the row of figures along the top. The complete absence of borders gives us no hint of the original size or configuration of this extraordinary fabric. This weaving is of such high quality that I was surprised to see both single and double wefts used with no discernable pattern. On one ex ample, narrow bands of single wefts lie along the edges of two ele ments of the design; all the other wefts are used in pairs, including those in the dark blue ground. A second example (Figure 13) shows the weft yarns alternating quite indiscriminately between singles and pairs, and here the ground has single wefts. It is difficult to understand the reasons for such changes. It may indicate that two weavers worked on this piece; perhaps one preferred the precision of single wefts, while the other liked the speed gained by using pairs. In any case, de spite the inconsistencies, this piece remains an extraordinary example of weaving, and must have been greatly treasured in the temple before it was destroyed. Decorative elements Two specimens in this category, both incorporating flax thread, are the two remaining exceptions to the exclusive use of cotton in the col lection of temple textiles. One of these specimens (Table 1, 16) is a row of connected tassels (each 2 by 0.6 cm) sewn to a narrow strip of 92 plain weave flax cloth. Ripped away from the rest of its fabric, this tassel border was found in the same area of temple fill as two large flax tassels (Table 1, 17), which are obviously a pair. The "head" of each tassel is 3.5 cm long, and is delineated by a band of decorative wrap ping 4 cm wide, composed of ten rows of alternating brown and white threads, centered by a pair of tan threads. The brown and white threads are of cotton, while the tan threads are of flax, as are the threads in the tassels. The technique of wrapping is a clever one which conceals the unused yams behind the visible ones until they are needed again. The "brown" cotton threads were originally red, as revealed in portions of the thread covered by other threads. Another tassel (Table 1,18) completes the set of decorative elements from the lemple collection. It is entirely of cotton except for the finial. In this interesting feature, the cords have been worked into a pompom, and a bit of flax cloth has been applied to hold the circular shape. The decorative pattern and the wrapping technique are the same as that of the large flax tassels discussed above. The red threads on this tassel and on the two large flax tassels are the only exception to the strict use of blues and natural color in the temple textiles described thus far. Textile miniatures. The largest group of textiles found in the temple, the woven minia tures, is comprised of forty-one specimens. They were found under rather different conditions from the fabrics which were part of the temple furnishings, discussed above. The miniatures were discovered in the temple sanctuary, bunched together in several groups. An un usual feature of the miniatures is that, although unfinished, most of them are complete as far as work had progressed, having a starting border, two side selvedges, and bare warp ends. Dimensions range between 14.5 and 61 cm long (excluding bare warps), and between 8 and 27 cm wide. They are technically similar to the previously dis cussed textiles: made of cotton, spun in the "s" direction, and with one exception, are limited in color to medium blue, dark blue, and the nat ural color of the cotton fiber. In addition to plain and tapestry weaves, one of the miniatures is woven with pile loops, a technique commonly 93 found among textiles of the Meroitic and immediate post-Meroitic pe riods. The textiles in this group were made on the warp-weighted loom as evidenced by the starting borders and the long warp ends to which loom weights had been originally attached. (A cache of some thirty loom weights was found in one of the temple crypts). The quality of these weavings is poor, in contrast to the earlier described groups. The yarn is unevenly spun, and weaving in many cases is very loose and open. These specimens were clearly not made by the same weavers who produced the furnishing fabrics of the temple. The smallest one (Table 1, 19; Figure 14, left) seems to have been a sampler ol tapestry-woven motifs arranged in horizontal bands. The figures include a three-lobed shape, ankhs, connected dots, connected diamonds, and a checkered arrangement. The yarn is coarse and lumpy; the quality of the weaving is competent, but entirely lacks the mastery shown by the temple furnishing fabrics. The remaining forty specimens have been designated as miniatures for want of a better name, since their function is not obvious. Twenty- four exhibit only the natural color of the undyed cotton and are woven in plain weave (Table 1, 20; Figure 14, right). The pile woven speci men is also undyed. Several pieces have narrow bands in blue near the top and bottom edges, and one has blue blobs at infrequent intervals worked in tapestry weave. Tapestry weave also was used to produce blue ankhs at regularly spaced intervals (Table 1, 21; Figure 15), and confronting snakes with three smaller ones below them (Table 1, 22; Figure 16). One miniature had an overall pattern of red and white checks - the only instance among the miniatures of the use of red-dyed yarn. An indication of the ultimate intent of the weavers is provided by two specimens that are more nearly finished than the other miniatures. The lower edge of each has a border of wrapped openwork with full fringe tassels attached. Furthermore, the starting borders of the upper edges had been removed, and decorative borders had been started across the top. In the smaller specimen (Table 1, 23; Figure 17) the border was simply a row of little balls made of the warps wrapped 94 with blue cotton yarn. This piece seems to be a miniature of one of the temple furnishing fragments (Table 1, 12; Figure 11). The top of the latter is missing, but these two specimens share in common an undyed ground weave, a tapestry-woven band near the lower edge using two shades of blue and white in very similar motifs, and a simple openwork border with thick fringe tassels. The larger almost-finished miniature (Table 1, 24; Figure 18) seems to have been a smaller version of the blue curtains that were described above (Table 1, 1,2,3; Figures 4, 5, 6). It is plain weave, dyed blue after weaving. Undyed yarns added to the lower edge have been worked into an elaborate border of wrapped openwork with thick fringe. Strung on one of the fringe yarns are three glass beads whose colors are dark blue, pale blue, and yellow. Nine tiny (5 mm) balls of undyed flax yam are spaced at regular intervals in a row immediately below the openwork border. These represent the only use of flax among the miniatures. The starting border across the top had been removed, and in its place a pattern of wrapped openwork was begun in blue. It extended for approximately 9 cm in various stages of comple tion. The woven miniatures are among the most puzzling finds from the temple. There is little doubt that they were in the process of produc tion; many had balls of weft yarn still attached, and the long warp ends would have been incorporated in some type of ending border. It is difficult to escape the supposition that they were being made in or very near the temple. Found in the temple sanctuary along with the minia tures were six balls of spun cotton yarn, containing many meters of yam ready to be woven. The previously mentioned loom weights re covered from one of the sub-floor crypts are further evidence. A weaving comb, used for beating up the weft, was found in the fill. While none of this is conclusive, the evidence does seem to point to weaving activity going on in the temple. Recovered from various rooms of the temple were nine non-woven miniatures; these are examples composed solely of wrapped openwork and fringe. Technically, they are like the openwork borders of the miniatures just described, and those of the temple furnishings, but dif fer in that no weaving is involved. They are simply patterned open- 95 work created by wrapping z-plied yarns around selected other plies, with fringe tassels tied to the lower edges. Each one is a complete, finished construction. Some of the pieces are simple and appear to be the work of a beginner. Two specimens, of undyed plies only, are technically complex (Table 1, 25; Figure 19). Two others are elabo rated with the addition of tiny blue yarn balls (Table 1, 26); on one they are placed in a row across the top, and the other has balls tied to the fringe yarns (Figure 20). The four just described show technical prowess and delightful imagination, but it is difficult to guess their purpose. One is tempted to suggest that they were patterns to be fol lowed by novices, or that they were practice pieces made by more ad vanced students. The final piece to be considered (Table 1, 27; Figure 21) fits into none of the earlier mentioned groups: it had not been used as furnishing fabric, nor can it be accurately termed a miniature. The two frag ments comprising this piece were found in the inner hall of the temple. Its unused condition, the quality of the yarn and the weaving, mark it as the work of the same novice weavers who produced the woven miniatures. But despite its fragmentary state, it was originally too large to be classed with them. A portion of the starting border and part of one selvedge have been preserved. An interesting technical feature of this specimen occurs just below the starting border. The first 5 cm of the warps have been left intentionally bare. This leads to the possibility, admittedly specu lation, that the weaver intended to remove the starting border and work a decorative border, as was seen in number 23 and number 24. However, the tapestry-woven motif is the most striking feature of this piece. It is a depiction of an offering table with a horned disc placed on the side opposite the spout. The same symbol was found on other ob jects from the temple including a ceramic basin and a stone offering table. This motif has been found also at other archaeological sites, and during the late classical period was associated throughout Nubia with the goddess Isis. 96 Conclusions. Since so many of the temples of the Nile Valley were converted into churches after the coming of Christianity, most of the sacred objects of the temples were inevitably lost. The discovery of furnishings or rit ual paraphernalia within the walls is extremely rare. As far as I know, our discovery of textile furnishings in the temple where they were once in use is quite simply unique. Despite the extraordinary circumstances of their discovery, the textiles themselves fit within the generally recognized parameters of textiles of late classical times in Nubia. These include the predominant use of cotton, all spinning in the "s" direction, weaves limited to plain weave including half-basket and basket weaves, tapestry weave, and weft loop pile. The use of color was for decoration and was generally limited to shades of blue and very occasionally, red. Other characteristics in clude the use of embroidered decoration, and the ingenious wrapped openwork borders. Then what has this discovery added to our knowledge of late classical Nubian textiles? The three blue curtains, all dyed in the piece, are a real departure from the prevailing practice. In analysing more than three-thousand textile fragments of this period from elsewhere on the site, fewer than ten have been s-spun cotton, piece-dyed blue. It now seems possible that these few were also pieces of the destroyed cur tains, since none of them was complete. A second contribution is the complex pattern of the bordered cloths. Nothing similar to these, with their bordered, elaborate, tapestry-wo ven rectangles, surrounded by an undyed plain ground weave has ever been reported. With the exception of the ankhs on the smallest cloth, the motifs are geometric, not iconographic. Yet, given their prove nience, the religious nature of these fabrics must now be taken into ac count when temple furnishings are considered. The outstanding quality of the tapestry-woven furnishings raises our estimation of weaving to higher levels of artistic achievement in this period. We had been aware that these people were competent spinners and weavers and were very skilled at sewing and embroidery. Frag- 97 ments of tapestry weaving had been recovered previously, but nothing so complex or so finely woven as these. The question arises, where were these textiles produced? We shall probably never be certain. Textiles are so easily transportable; these may have come long distances. A possible origin may have been the city of Meroe itself. Because of the climate, few textiles have been re covered from there. But, one small fragment of wrapped openwork has been reported (Griffith and Crowfoot 1934: 10, Figure 1) and it is far finer than any from Qasr Ibrim. Grace Crowfoot's descriptions of the several textiles she examined, and her thread counts, indicate that yam and weaving was extremely fine at Meroe. However, there is no doubt that weaving was done at Qasr Ibrim. The miniatures may not have the quality of the temple furnishings, but their discovery, in their unfinished state, surely provides a rare view of textile production of this period. The miniatures pose as many questions as they answer. Were they intended for eventual use, and if so, what was it? What is the relationship between the furnishing fab rics and the miniatures? Similarities of motifs and structural parallels between the two indicate that some sort of relationship existed. The answers to these and other questions must await further excavation, hopefully in the near future. 98 Footnotes 1. The Egypt Exploration Society of London has been excavating Qasr Ibrim since 1961 under license granted by the Egyptian An tiquities Organization. Reports of the excavations may be found in the Journal of Egyptian Archaeology beginning with Volume 50, 1964. All of the textiles mentioned in this article are the property of the Egyptian Antiquities Organization and are housed in the Cairo Museum. I would like to express my appreciation to the Society for its sup port of the Qasr Ibrim Expedition these twenty-five years. Thanks for financial and other support are due also to the Smith- sonian Foreign Currency Program, the American Research Cen ter in Egypt, the University of Kentucky, and the Egyptian Antiquities Organization. Among the individuals who have con tributed in various ways to this work I would like to gratefully acknowledge the members of the 1986 expedition; Messrs. Abdin Siam, Magdi Abdin and Abdul Hakeem of the Egyptian Antiqui ties Organization; Elisabeth Crowfoot, Janice Yellin, Robert Bianchi, LoErna Simpson, Jean Hoskin, and most especially my husband, William Y. Adams. 2. For a fuller description of the temple and its contents see: Journal of Egyptian Archaeology, Vol. 73, 1987 (in press); andArcheo- logie du Nil Moyen , Vol. 3, 1988 (in press). 3. Archaeological textiles cannot be considered white by today's standards. Many factors affect the appearance of colors which have been buried. For these fabrics, the conditions of use, rather than the actual age seems to have determined their present appearance. Those textiles which furnished the temple were darkened with smoke from lamps and incense. The oily residue left by the smoke tended to attract dust, which further darkened the colors. The appearance of dark blue was affected least by these factors; undyed yarns were changed the most. The pale blues and medium blues have taken on a greenish hue, a combina tion of the blue and the soiling elements, which by themselves leave a tannish or golden color. By contrast, the textile minia- 99 tures, which were abandoned before they were finished, were never put into use. Their blues are still quite bright, and some of the undyed areas are much closer to the original almost-white of the unspun cotton. (Unspun cotton still in its calyx has been found from this period at Qasr Ibrim.) In describing patterns with complex color relationships I will use the term white for the sake of simplicity; areas of plain weave will be called undyed. 4. The wrapped openwork technique has not been reported from Egypt, but seems to have originated some three-hundred miles up the Nile in the Sudan. From Kerma, several specimens were found which are dated to roughly 1500 B.C. They are now in the Museum of Fine Arts, Boston. There is a hiatus of about a mille- nium and a half between the Kerma openwork and that of late classical times in Nubia, but knowledge of its production might have been kept alive, just as certain basketry techniques have been practiced from the time of the pyramids right through to the pre sent day. 5. The dimensions given in Table 1 for the bordered cloths are ac tual measurements of each specimen as it appears today in its fragmentary, but reassembled state. However, dimensions of the patterned portion of each specimen can be computed by adding in the missing parts. Hypothetical overall dimensions of one of the cloths and the width of another can be calculated by utilizing in formation gained from the presence of selvedges and end borders, and by assuming that the patterned portions were centered on the ground weave. The figures are as follows: Specimen Dimensions Overall size of patterned area small bordered cloth 13 by 18 cm unknown medium bordered cloth 16 by 32 cm ? by 50 cm large bordered cloth 46 by 73 cm 74 by 99 cm. 6. This motif is also found painted on pottery and molded in faience from several Meroitic sites, and is associated with the Isis cult. 100 Bibliography 1. Elisabeth Crowfoot, Openwork fringes from Qasr Ibrim, Meroitic Newsletter No. 23, Paris, 1984. n.d. Notes on the Nubian Textiles from Gebel Adda, Manuscript on file in the Egyptian Department of the Royal Ontario Museum, Toronto. 2. F.L. Griffith and Mrs. G.M. Crowfoot, On the early use of cotton in the Nile Valley, Journal of Egyptian Archaeology Vol. XX, Parts I and II, 1934, 5-12. 3. D. Randall-Maclver and C.L. Woolley, Karanog the Romano- Nubian Cemetery (Eckley B. Coxe Junior Expedition to Nubia) Vol. IV, Philadelphia: University of Pennsylvania, Egyptian De partment of the University Museum, 1910. 4. Christa C. M. Thurman, Ancient Textiles from Nubia, Chicago: The Art Institute of Chicago, 1979. 5. Kurt Weitzmann, Byzantine and early medieval antiquities in the Dumbarton Oaks collection, in Ivories and Steatites Vol. Ill, Washington, D.C.: The Dumber-ton Oaks Research Library and Collections, 1972. 6. R.E. Witt, Isis in the Graeco-Roman World, Ithaca: Cornell Uni versity Press, 1971. Museum of Anthropology University of Kentucky 101 TABLE 1 E6YPTIAN ANTIQUITIES ORGANIZATION QASRIBRIM REGISTRATION OBJECT DIMENSIONS NUMBER NUMBER 1. Curtain 74x81 cm. 86T/025 QI86R660 2. Curtain 85 x 115cm. 86T/023 QI86R661 3. Curtain 42 x 56cm. 86T/096 QI86R654 4. Case 14 x 18 x 7cm. QI85. 12. 19/25 QI86R70 5. Long bag 91 x30cm. 86T/166 QI86R643 6. Leather container 31 x 14cm. 86T/032 QI86R71 7. Bordered cloth 31 x 38cm. QI.85.12.19/46 QI86R649 8. Bordered cloth 2 1 x 48 cm. 01.85.12.19/42 QI86R656 9. Bordered cloth 61 x91 cm. QI85. 12. 19.43 QI86R659 lO.Fragmentwlth 3x27.5 cm. QI85. 12. 19/45 QI86R653 ankhs 2x 17.5cm. 11. Ladles 1 8 x 8 cm. Ql 85. 12. 19/41 QI86R652 1 2.Border with 5x34 cm. QI85. 12. 19/47 QI86R650 fringe fringe-Bern. 1 3. Border 2.5 x 10cm. 86TA089) QI86R657 1 4.Border 7.5x2.8 cm. Ql 85. 12. 19/44 QI86R648 IS.Meroiticsa 64 X 1 2 cm. approx. QI85. 12. 19/39 QI86R662 1 S.Tassel border 3 x 33 cm. 86T/086 QI86R644 171arge tassels 16x2.2 cm. 86T/087 QI86R645 1 S.Cotton tassel 16.5x 1.5cm. 86T/140 QI86R651 1 9.Tapestry sampler 1 4.5 x 8 cm. 86T/294 QI86R639 20 Unfinished 52 x 12cm. 86T/293 QI86R639 miniature bare warps 28 cm. 21 Miniature with 34 x 22 cm. 86T/054 QI86R640 ankhs bare warps 5, 21 22. Miniature with 32x21 cm. 86T/044 QI86R638 snakes 23. Small miniature 27 x 14cm. 86T/055 QI86R640 24.Blue miniature 60 x 25 cm. 86T/031 QI86R658 25. Openwork 4.5 x 7 cm. 86T/070 QI86R641 miniature fringe- 12cm. ,26.0penwork 4x5.5 cm. 86T/139 QI86R641 miniature fringe - 8 27. Fragments with 1 6 x 44.5 cm. 86T077 QI86R600 offering table Note: As the fiber, direction of twist. weave and colors for each specimen has been covered In general remarks or have been mentioned specifically in individual descriptions, they have been omitted here to avoid duplication. All measurements give the dimension of the warp direction first, the weft dimension second 102 WARP COUNT WEFT COUNT TAPESTRY JOIN TYPES 14-15 12-14 7-8 7-9 9-11 11-13 10 11 6.5 pairs 26 singles, 1 1 pairs slit 18 10 12-14 32-38 singles, 12 pairs slit, dovetail, interlocking 11-17 42-43 singles, 16 pairs dovetail .interlocking 9-12 40-41 singles, 27 pairs dovetail, interlocking 10 36 slit, interlocking 14 41 slit 12-14 32- 38 singles, 12 pairs slit, interlocking 13 40 slit.inter locking 13 31 interlocking 11 -12 pairs (49-50 blue, 24-38 whi interlocking (32-42 blue, 24-30 white pairs 11 plain weave, 20 tapestry weave not recorded 6 9 plain weave, 16 tapestry weave slit, dovetail 4, 17 (tap.wv.) 5 plain weave, 20 tapestry weave slit 7 1 1 plain weave, 30 tapestry weave interlocking 8 12 7-8 1-12 plain wv., 17 tapestry wv. dovetail 103 Figure 1 Map showing the location of Qasr Ibrim 104 Figure 2 Photo of Qasr Ibrim a. isi m ais I I 1I I' Figure 3 Sketch-plan of the temple 106 Figure 4 Number 1, curtain III lilli Figure 5 Number 3, curtain with openwork border Figure 6 Detail of border on curtain Figure 7 Number 4, double-chambered case 110 Figure 8 Number 7, small bordered cloth Figure 9 Number 9, large bordered cloth Figure 10 Close-up of large bordered cloth Figure 11 Number 12, fragments of tapestry-woven border with openwork Figure 12 Number 15, reassembled fragments with Meroitic sa 115 ;r -: m " - *=•" **>>* s.^m':mF^^ .., •-:• . „ ^ii^ffc?. 'fcfe- „.- „ aaes * „ sii«^^^. ,* Figure 13 Detail of number 15 showing random usage of single and double wefts 116 w^j?***- <:k ->-%d- - x .- •&& -^m ,. J Figure 14 Left, number 19, sampler; right, number 20, plain weave miniature 117 I IX f > i Figure 15 Number 21, miniature with ankhs 118 Figure 16 Number 22, miniature with snakes 119 Figure 17 Number 23, small almost-finished miniature 120 Figure 18 Number 24, blue-dyed almost finished miniature "curtain" 121 Figure 19 Number 25, undyed openwork 122 Figure 20 Number 26, openwork with blue balls 123 Figure 21 Number 27, offering table with horned disc