The present carpet of lattice design and with marvellous colours belongs to a small and rather rare group of long narrow carpets which were made in villages in an area in East Azerbaijan and which survived from the 19th century.
An almost identical carpet to our example is published by Eberhart Herrmann, Asiatische Teppich- und Textilkunst, vol.4, Munich, 1992, pl.64, pp.140/141, which differs only in a couple of minor details and in the colour and design of the guard stripes. The remarkable similarity of the two pieces indicates that they were made in the same village or workshop. A third example on ivory ground and dated AH 1222/1807 AD, was previously in the Vojtech Blau Collection and sold at Sotheby's New York, 14 December 2006, lot 28. It has the same basic design although there are obvious differences.
The lattice field design in the present lot can be found in Persian, Turkish and Caucasian carpets. As a favourite and most saleable pattern in this version it was brought back to Persia from India in the 18th century. Some of the finest variations ultimately derive from the Kirman 'vase' carpets. The Textile Museum in Washington has a fragment of a Kirman 'vase' carpet which clearly shows an overall lattice design formed by serrated leaves each containing four different flowerheads which are surrounded by most finely drawn flowering vine (May H. Beattie Carpets of Central Persia, exhibition catalogue, Sheffield and Birmingham, 1976, pp.75-76, pl.10 and 49). Later in date and with a rather basic design is an example illustrated in Werner Grote-Hasenbalg Der Orientteppich. Seine Geschichte und seine Kultur, vol.III, Berlin 1922, pl.64. (vol.1, pp.133-134) somewhat strangely attributed to Sultanabad. A further possible influence is shown in the later Caucasian Dragon carpets, for instance the Dragon carpet in the Old Mosque, Edirne (Serare Yetkin, Early Caucasian Carpets in Turkey, vol.1, London 1978, pl.8). A further development of the pattern is shown in a well-known group of Karabagh carpets where the "leaves" join up to became a homogeneous field around lozenge medallions.
How the original design changes can be further seen in the Karabagh carpet illustrated in Peter Bausback, Alte und Antike Orientalische Knüpfkunst,", exhibition catalogue, October 1983, Mannheim 1983, pp, where the leaves are no longer indicating a specific direction which they still clearly do in our example; or in an even more simplified version in another Karabagh kelleh sold with Rippon Boswell, Wiesbaden, 10 November 1990, cat.32, lot 193, which retains the flowerhead surrounded by four stylised flowers and set into a fine lozenge, but where the leaves which form the lattice are merged.
Another theory regarding this specific type of the present North West Persian carpet has been presented by Raoul Tschebull. He thinks that that very area in Azerbijan became a cheap and most productive place for trade men mainly from Tabriz, an already established centre of carpet production, who were commissioning carpets for the growing carpet market in the Western world by the end of the 19th/beginning of the 20th century. Known today as "Heriz" named after the largest town in this area, these carpets show a mixture of new designs given from Tabriz merchants and the traditional colours produced by the weavers in the villages. The size of the carpets, too, was according to the demand of the Western customers much smaller than those being woven for the own use of the nationals (Raoul E. Tschebull, "Heriz. A Historical Perspective", Hali, January 1997, issue 90, pp.64-66). The group of kellehs such as the present lot are assumed having been woven during the very beginning of this period but heavily reflecting features of the traditional weaving than the new one. They are of the traditional size and colouring, but the structure is the same as in export carpets before 1920. It also appears that the design has been created rather by the village people as it barely can be found in later (export) Heriz carpets (op. cit).