Twenty-five years after Martin discovered the Konya carpets, more Seljuk carpets were found in the same way in the Esrefoglu Mosque
In Beysehir, a district capital lying on the shores of the Beyşehir Lake. They were discovered by R. Riefstahl and brought to the Mevlana Museum, Konya. Three of these carpets, which all bear the same characteristics as the Konya carpets, are worth examining closely. In fact one fragment, of which only a small part of the field border remains, is very similar in composition particularly to one of the Konya carpets. The fourth piece is from the 15th century. The first fragment measures 1.70x2.54 m and, on a dark blue field exhibits a composition of a series of diagonal rows of light blue lozenges and a pair of hooks extending from both sides to form a small diamond which is dotted with red or yellow fillings (Pls. 12, 13; D. 12). The main lozenges contain eight-pointed stars. They are either light yellow or red with dark blue outlines, or they are dark blue with yellow or red rectangles in the center. The border, of which only one small fragment remains, must have consisted of a Kufic motif in white on a red field.
The field pattern of this carpet with hooked lozenges is used as an emblem on the tea boxes of the Ceylon Tea Center near Piccadilly in London. Buddhism which disappeared in India, the country where it was born, lives on in Ceylon. It might be a Buddhist symbol that came to Anatolia from the Uighurs. The same motif later appears in the 18th century Turkish carpets and prayer rugs. We see a characteristic example of this in the borders of the Ladik prayer rugs with three mihrabs divided by thin columns (Ills. 3, 4).The second fragment measuring 1.16x0.49 m (Inv. No. 867), is a very damaged border piece of a woollen carpet which has lost its pile. It is a 13th century Anatolian carpet. The composition is schematically identical to that of the second Konya carpet, although here the motif is brown on light red; the Konya carpet is light blue on blue (Pl. 11; D. 11).
The third example consists of fragments of a carpet which presumably was originally five meters long. In an article by Riefstahl on the subject, he refers to it as a Selçuk floral carpet, approximately five meters in length, with designs reminiscent of those seen in the stonework on the left side of the main façade of the Turumtay tomb in Amasya, and also on the Gök Medrese and Çifte Minareli Medrese, both in Sivas. Riefstahl claims that the, hooked lozenges of this carpet were taken from Byzantine textiles. However here the system of devices is quite different and has become very Turkish in character. (3)The carpet in question contains an endless repetition of geometric floral motifs arranged in rows a reversed flower bud projects from a vertical stem and from its tip extends a hooked palmette facing right or left in alternate rows. The palmette itself bears a forked band with hooked leafy tips. The base of each stem joins the tip of another in the next row.
The main border is covered with a repeated Kufic-like device in which a pair of hooked figures, "arms akimbo," alternate with two half figures of the same kind, each motif separated by double lines and obtuse triangles. The guard borders are flanked by bands of chain-like devices, and contain rows of vertical motifs identical to those found in the first Beyşehir carpet. The Beyşehir carpets were all brought, as mentioned, to the Mevlana Museum, but later one of them disappeared and hereby hangs a tale. In 1957 Erdmann was shown a photograph of a carpet fragment by the director of the Textile Museum, Washington, DC, one which he was about to purchase as a Selçuk carpet. When the director learned that it was a missing fragment from a Beyşehir carpet he withdrew his bid to purchase it. After an account of this was related in a Turkish publication in 1973, a rather large fragment of this Beyşehir carpet, one which had previously only been known from its illustration and description in Riefsthal's publication of it, was put up for auction. It was part of the estate of a Mr. Pozzi, the one-time French Ambassador to Turkey who had died in Paris. This fragment was purchased at that time by Dr. Edmund Unger for inclusion in the Keir Collection, London. (14) The carpet, formerly described in the literature about it as having a border with dark blue motifs on a light blue field, is now correctly seen in reality to have instead a light red border field with a dark blue motif, and one of the guard borders, which was thought to have had a light yellow field with reddish brown twin Kufic devices, instead has a light brown field with reddish brown motifs (Pl.14; D. 13). (5)
This Beyşehir fragment, which measures 2.07x1.85 m, has a 2 "S"-ply warp of undyed white and light brown wool, with a 2- and rarely a 3-ply weft of reddish brown wool. The knots are Turkish. The remaining mark lines on the incomplete borders give us a clue about the width of the field, which appears to have been approximately1.30m. The knots of the narrow border are visible in the upper part. The border, measuring 0.63 m, is half the width of the field, thus the carpet was 2.60 m in width when intact. This is one half of the 5 meters which was the previous estimate of its original length. None of the Selçuk carpets from Konya or Beyşehir has attained such extensive dimensions. This fragment, bearing as it does floral motifs, shows the development from totally geometric motifs prevalent in the other Selçuk carpets towards floriated motifs and may therefore be dated to the end of the 13th century. This is further corroborated by the fact that this date corresponds to that of the completion of the Eşrefoğlu Mosque, Beyşehir. () A few years after Dr. Unger's purchase there appeared in a private collection in Germany yet another fragment of similar size from this same Beyşehir carpet. This time, the mark lines on both the width and length of the border give the limits of the carpet in its complete form. The border actually exists only on one of the long sides of the carpet. The other three borders are missing. With this data in hand from this carpet fragment, we must assume that this third Beyşehir carpet was large and did originally measure nearly 5 m in length and 2.60 m in width thus verifying Riefstahl's stated estimates (Pl.15). But Beyşehir's claim to being part of the historical development of Selçuk carpet making is not only these four carpets found in the Eşrefoğlu Mosque. Seventeenth century kilims also found by R. Riefstahl in Beyşehir attest to the fact that more discoveries and research are necessary before we have the total picture of this area's contribution.