The design builds on seven large simplified palmettes along the central axis, a design derived from sixteenth and seventeenth century Persian court carpets. Half palmettes toward the edge of the field indicate an endless repeat design. The palmettes are filled with various secondary motives: branches, abstract flowers with a chessboard design, animal forms(?) and botehs. Caucasian Kuba rugs also show similar palmette motives.
Click on image to view in larger size. View rug in extra large size (jpeg opens in browser window).
The row of palmettes is flanked by claw-like floral forms linked b dotted bands (stems?) and other scattered motifs, many of which are rendered with red or rose outlines, like botehs, diamonds, and smaller buds. A rather unusual feature is the use of two white-ground stylized dragons-in-tow borders spearated by a narrower nicely abrashed mahogany-ground middle border carrying in turn botehs and simple rosettes.
While this is certainly not a tribal rug (for a start, it is wider than the common portable loom used by nomads), it shows the charming simplifications of city and court designs that we see in many village rugs in the Hamadan area.
The rug measures 7.8 x 4.7 foot (238 x 145 cm) and is quite heavy and substantial. The warps are white cotton with no displacement, the wefts are light brown wool (looks like two thin shots in the same shed). The knots are symmetric (Turkish knots) , 28 / dm horizontal and 44 dm vertical (ca 7 x 11 kpi = 77 kpsi). The top has a brown kilim of about one inch, at the bottom it may have been slightly narrower, not so much of it remains. 10 warp ends were bundled in one knot. What seems unusual is that the rug seems to have been woven upside down (in relation of the direction of the palmettes / flowers) for whatever reason -- or, may be, the Hamadan weave technique leads to the pile pointing upwards?
The dyes are saturated, the colours look very fresh, contrasty and well-balanced. There is no tip fading that would be typical for many early synthetic azo dyes. The exception is a Fuchsin-type dye used in the borders (about Fuchsin, read below in section on attribution).
The field colour is a very dark indigo blue with a nice abrash. In the palmette motives, the medium-to-light indigo blue is matched by a saturated probably madder-based mahogany red. This is complemented by medium green in the red palmette triple outlines and many filler motives, and a very light yellow used in a few places. The claw-like vegetal forms use several slightly abrashed shades of salmon, from a deep and intense coral red to a light rose colour. Whether the coral to rose red is natural, synthetic or even a result of overdying natural with synthetic I cannot establish. Madder dyes with an alum mordant can result in very vivid shades of orange-red that many mistake for synthetic dyes. On the other hand, synthetic dyes were available around the turn of the century, especially at the mor ecommercial production centres, and the coral red is really quite strong. I am just not sure. There is no bleeding I can see, not the slightest tip fading, and the "damp cloth" test indicates that the colours are fast - no trace of colour left on the white cloth after wiping.
The overall condition is good, with some minor problems that can be quite easily corrected. The construction is very solid, the pile is dense and hard-wearing. The rug lies flat on the floor. The pile is remarkably high throughout (it has never been very long from the start, perhaps 7-10 mm), the wool is soft and glossy. There are actually few signs of wear from use, ecxept of course for the ends and the selvages, which have at some point been sheathed with a dark brown woolen cord (looks like a rather old repair) and have come loose in many places. Also, a thin band of cotton or linen cloth has been sown underneath along the sides to secure the selvages (can be seen in the photo of the back). The selvages should better be secured / rewrapped on both sides.
While the rug is not very dirty and does not smell, it is dusty, especially on the back, and should perhaps best be washed. I can also see no stains. I can spot three small damages, a very small area at the end of the left branch inside the forth central palmette (counted from the bottom), and two small holes both less than a square inch) that have fixed through darning, one in the third central palmette, another near the stem of the red 'claw' to the right of the second mahogany palmette. The end kilims are partly there, even with some of the decoration line, and some knotted fringe remains. The ends should also be tidied up.
(1) The color combination and the bold forms of the palmettes and leaves that remind me of the drawing in the Songur/Kolyai and Sauj Bulaq areas could indicate this is Kurdish work. The dragon repeat border, the rug's structural features (no warp depression, cotton warps and ligh brown single wool wefts) as well as the overall palette show a strong resemblance to a rug No. 26 Qombâd shown in Willborg's book catalogue "Hamadan", an important publication on weavings from the Hamadan district. The weave pattern (the way the rug looks and feels on the back) showing the typical Hamadan basket weave but in a somewhat irregular pattern of warp spots showing, is also very close to the rug Willborg attributed to Qombâd, a village about 20 km south-east of the city of Hamadan. The rule described by Willborg for Qombâd, "one main border (...) flanked by three narrow guardborders on each side" is obvioulsy not followed in this rug. It is also considerably larger than the Willborg example. Other provenances within the Hamadan district are therefore feasible. The weave itself is distinctly Hamadan.
The age of this rug is difficult to determine. The colours give a clue, though. The dragon border uses Fuchsin in small amounts. Fuchsin is an early synthetic mauve dye that fades to a greyish tone. It was used mainly in the last quarter of the 19. century; its use ebbed around the turn of the centrury because better dyes became available. Note however that this is no evidence tha this rug is pre-1900, though. Fuchsin may still have been in use well into the 20. century.
(2) I think I have to revise my attribution to the Hamadan area since the rug has double wefts (in the same shed), not the single weft that characterises Hamadan weaves. (Thanks to my Swedish friends for making me aware of this.) The weave looked similar enough but this detail escaped me. Therefore, the rug could be north-west Persian or even Caucasian. People have suggested it could be a Cuba rug, but the warps are not depressed and not wool as would be typical for Kuba rugs; Karadja area, but the rug has double, not single wefts. Baku? perhaps the wrong palette for Baku. Karadagh? Karabagh? I have to admit I just don't know, it remains a mystery rug. I have requested help on attribution and will update you on the results, if any experts come forward.
(3) In response to a request for attribution help on the Orientalrug yahoo mailing list, there has been no complete convergence of views. Views range from a suggested Karaja provenance (north-west Persian, equally single-wefted) to a confirmation of a wider Hamadan origin by Barry o'Connell, list owner, who thinks even double wefts, as long as they are same shed, qualify technically as single-wefted, and suggests it may have been woven by descendants of refugees from the Persian Caucasus displaced by the Russians in the 19th century. Another list member believes it to be "a Hamadan woven copy of a Caucasian sumac". It just shows that things often aren't easy in rugdom! Going back to Willborg's Hamadan book, there are indeed some rugs here that have two woolen wefts in the same shed: no. 3 Hamadan Province (dated 1825-1875) in the first 20 cm, no. 4 Hamadan? (dated 1770-1830) and no. 9 Gharaghan (dated 1910-1925), with two weft shots in the same shed, albeit cotton wefts, not wool as in this rug, and no. 24 Mehraban (dated 1910-1930) has sometimes one cotton and one wool thread in the same shed, etc. etc. At least this shows that the general rule that Hamadans have just one weft thread after each row of knots has exceptions, particularly in the older pieces. In general the use of wool for the weft seems to get rarer in the more recent Hamadan rugs (first third of 20. century), though to me this seems insufficient to firmly establish a pre- 1900 date.