About the Antique Rugs of the Future Project

Sheep Breeds of Azerbaijan

Sorting, Washing, Carding, Spinning

"The advantages of handspun yarn to machine spun yarn"

Rediscovery of Ancient Natural Dyes
Our Natural Dyestuffs


Difference between synthetically and naturally dyed rugs

Weaving and Finishing Steps

Galleries of ARFP Caucasian Azerbaijani Rugs

back to "Early Karabagh rugs and carpets" main page


AH 1001 (1592 AD) dated Dragon Rug with Balanced Design. Karabagh. Textile Museum R 36.1.4. (formerly R 2.47), formerly George Hewitt Myers collection. . late 17th century would be a more plausible date for this rug.

A carpet with many puzzling features. The field is a dark, rather brownish brick red; the band systems blue-green and yellow-orange, enclosing a light rose red diamond area about the centerpiece; the border a dark and glowing medium blue. The shades of color do not match those of normal dragon rugs, nor does the weave, which is more flexible and which lacks the heavy, cabled single weft shots which are so characteristic of the weaving habits which prevailed where most. of the antique Caucasian carpets were produced.

In this carpet the forms of almost all of the palmettes are decidedly degenerate; the band systems simplified and distorted in their drafting so that they become a mere grating in the upper part of the field. The octagram centerpiece, with its muddled decor, lends a jarring note, as do the four "signboards", with two chickens on each one.

The dragons, on the other hand, come closer to Chinese and Persian models than in the best of the 17th century dragon rugs and the flaming lions are clearly recognizable, which could not be said of those in the early carpets. So also are the ducks more clearly identified. At the base of the field are two cranes, based on Chinese models but here corrupt, a symbol of rank among other concepts. At the head of the carpet is an inscription which has been translated as "On the date of the fifth of the sacred Muharram of the year 1001. Husayn Bek, his servant." This would have been Oct. 13, 1592. However the date is not expressed correctly and it has been also read as "exactly Nov. 15, 1592" or as A.H. 1101 (1689 A.D.) and Oct. 17, 1089.

A completely centralized design for a dragon rug seems strange for the late 17th century, much more so for the late 16th, yet such things are possible. The four tipsy, masquerade party steinbocks do not occur in other dragon rugs, except the Graf carpet, which may have been a 19th century copy in view of the degree of degeneracy of its details. The balance of the animals are poorly drawn, beginning with the chilins and the pheasants, which resemble gallinules. The "chi" are poor, and in the top of the carpet, forgotten.

It is not really a pleasure to destroy old idols but in any serious art study the application of new perspectives must at times produce that result. It seems probable that the present carpet represents a free copy, made centuries later, on the model of Husayn Bek's carpet by a weaver whose imitative skills found better accomplishment in dealing with several animal forms than with the other motifs. The octagram centerpiece may have been her invention, replacing the "blazon" palmette which the two dragons should have been protecting. Illiterate herself, she may have made the errors in transcription that have given us such a variety of possible datings—applicable to the model. The weaver may have been working somewhere in the Karabagh province, but not near the original area of production.

The cranes at the base of the field also occur in Brooklyn Museum 65.250; in the Jackling dragon rug, 251.59 of the California Palace of the Legion of Honor in San Francisco; in the Graf carpet, now decimated, in East Berlin: in the former Stefan Bardini dragon carpet and in several others. They are usually taken for ducks, misunderstanding having so greatly shortened the legs. The corresponding Chinese version may be seen in a rug on loan to the Brooklyn Museum from the Krickson Collection (Ellis I96S: p. 40).

Size: L. 4.45 m. (14'7") x W. 1.83 m. (6'0").
Warp: Z2S wool in a light natural range with occasional dark fibers. One level.
Weft: Z2S wool in natural mixtures, dyed a good red. Two shots, sometimes three.
Pile: 2Z wool. Gordes knotted, pile lies straight. 8 horiz. x 10 1/2 vert. per in. (85 per sq.in.).
Sides: Cut. Ends: Cut.
Colors: Ivory; black-brown; brown: yellow-orange; yellow: dark brick and light rose reds; dark medium and fight medium blues; blue-green: light yellow-green; dull light violet.
Condition: Some areas rewoven and a great deal of re-knotting, now faded. Outer guard lost. Wear has been especially severe along the center line
Published: Pope 1925: III. 2; Bode and Kühnel 1955, 1958; Fig. 40; Erdmann 1962: Fig. 101: Ellis 1970: Fig. A.

Lit: Charles Grant Ellis "Early Caucasian Rugs", plate 11

The inscription at the top of the field: "On the date of the fifth of the sacred Muharram of the year 1001 (1689). Husayn Bek, his servant."