This saf is very closely related to a small group of examples with field of varying colours and differing designs which have been published. One of these, comprising seven niches, is in the Textile Museum, Washington (Prayer Rugs, Textile Museum, Washington, D.C., 1974, XXIV, p.80-81). A second with six niches is in the Marshall and Marilyn Wolf Collection (Daniel Walker, Flowers Underfoot, Indian Carpets of the Mughal Era, exhibition catalogue, The Metropolitan Museum of Art, New York, 1997, no.38, figs 132 and 133, pp.136-7). A third, comprising six niches arranged in two rows of three, was formerly in the Gueron Collection and sold in Italy in 1993 (U. Sorgato advertisement, Hali, Aug/Sept. 1993, pp.58-9). Two further examples were in the Untermyer Collection, sold Parke-Bernet, New York, 10th and 11th May 1940, lots 210 and 211, and there is a closely related cotton single niche in Philadelphia (Charles Grant Ellis, Oriental carpets in the Philadelphia Museum of Art, Philadelphia, 1988, no.66, pp.237-239). Further examples are noted by Ellis as being in the Perez Collection (two rows of three niches), in storage at the National Museum of Art, Lisbon (various fragments including much of five niches), and in the Reves Collection at the Dallas Museum (3 niches). A further example was offered in these Rooms, 14 October 2004, lot 25. All examples have had a considerable amount of wear and sometimes damage, have repiling in areas, and are in part composite from the same original. The design of ours is very close to the cotton example in Philadelphia discussed by Grant Elis, but this example does not use any cotton at all in the pile, and the drawing is better.
In his discussion of the group, Daniel Walker uses their similarities to three rugs and fragments to attribute them to the Deccan, and almost certainly Warangal (Walker, op.cit, pp.133-4). He also shows a group of similar prayer aches carved into the courtyard of the Jami Mosque in Hyderabad, completed in 1597. The three woven comparisions are all in the storerooms of the Victoria and Albert Museum; one was said to have come from the collection of the Nizam of Hyderabad while the other two were thought to have originated in Warangal. One of these appears to be the same as that published by Steven Cohen in his article which first suggested Warangal as the origin of the group (Stephen Cohen, "Textiles", in G. Michell (ed.): Islamic Heritage of the Deccan, 1986, pl.10, p.122).