Regarded as a whole, the borders of Asia Minor rugs show but slight relationship to either the geometric patterns of the Caucasian, or the floral patterns of the Persian; for as a rule the geometric features either are subordinate or suggest an origin by degradation from floral designs, and the floral features generally are represented by an orderly arrangement of disjunct forms rather than by continuous vines with pendent flowers. Yet there is no doubt that some were copied by Caucasian weavers, and that many were derived from Persian patterns. They are frequently, however, more artistic than the former, more interesting than the latter, and rival both in beautiful colouring and delicate drawing.
Primary Stripes.—Several different stripes peculiar to Ghiordes prayer rugs are illustrated in Plate [G], Figs. 1, 2, 3, 4, and 5 (opp. Page 192). The first of these is probably the oldest. It is so strikingly suggestive of the Herati design of rosette and attendant leaves that there can be little doubt of its Persian origin. The seed-like processes of the alternate rosettes are noticeable. The second, which is found in many of the Ghiordes rugs, shows the same pattern more conventionalised, with the rosette resembling an open pod and with the leaves almost octagonal-shaped. The third is a still greater evolution of the same pattern in which leaves and rosettes of nearly equal shape and size are placed at three angles of a quadrangular space. The last, which is a very elaborate pattern somewhat similar to the first, is found in a few old rugs. Besides these, a number of parallel lines similar to those of Kulah rugs (Plate [G], Fig. 12) are sometimes seen in the Ghiordes.
Plate G. Primary Border-Stripes of Asia Minor Rugs
One of the best known Ghiordes stripes found in odjaliks and sedjadehs is shown in Plate [G], Fig. 6. It is a broad stripe with wide, vine-like bands covered with rows of small flecks or flowers. Between each flexure of the band are designs probably symbolic of early sun worship.
In Plate [G], Fig. 7, is represented the most usual and beautiful stripe of the Ladik prayer rugs. The principal motives are Rhodian lilies, and rosettes identical with Persian forms, that are probably conventionalised roses. Another stripe, in which the lily is replaced by a conventionalised vine, is shown in Plate [G], Fig. 8. Both these stripes are found only in Ladik rugs. Other stripes seen now and then in Ladik and Melez sedjadehs are illustrated in Plate [G], Figs. 9, 10, and 11, each of which shows a conventionalised leaf.
One of the most typical Kulah stripes is seen in Plate [G], Fig. 12. It might in fact be regarded as a number of parallel stripes, but as will be evident by observing a large series of Kulahs it serves the function of a single broad stripe. Often the separate bands are replaced by a ground of uniform colour marked by parallel rows of minute flowers of regularly varying colour. Occasionally this stripe is copied by Ghiordes weavers.
In Plate [G], Fig. 13, is a well-known stripe that appears both in Kulah and Rhodian rugs. As previously explained, its origin is probably floral, though the drawing is geometric. A formal stripe found in Bergamos, and consisting largely of diamonds and eight-pointed stars, is represented in Plate [G], Fig. 14.