Resembling the last is the pattern shown in Plate [I], Fig. 14, which is commonly seen in Shirvan rugs and kilims. It is purely geometric and resembles one used in Western Asia Minor rugs of the XV Century, from which it may have been derived.
Though greatly conventionalised, the pattern represented in Plate [I], Fig. 15 is a vine and leaf derived from much more ornate forms, which may be seen in a XVI Century Asia Minor piece that is in the British Museum. It is now seldom copied, but was once a popular pattern for the Kazak and Kutais weavers.
Some form of the latch-hook appears in a large number of Caucasian stripes, but mostly in nomadic pieces. Figs. 16, 17, 18, 19, and 20 of Plate [I], represent patterns found mostly in old Kazaks and kindred rugs. The last one is also occasionally used as a secondary stripe. The patterns shown in Plate [I], Figs. 21 and 22 are from stripes sometimes seen in the Daghestan and Shirvan districts.
A much more interesting stripe because of its well authenticated antiquity, is the one shown in Plate [J], Fig. 1 (opp. Page 228). It is found in rugs made in Southern Caucasia two centuries ago, and according to Dr. Martin has been used since the XII Century. Probably as the result of copying, the design appears reversed in many old rugs.[30] These stripes are seen in comparatively recent pieces, but principally
in those of the Daghestan and Shirvan districts. Though the latch-hook is suggested by the small triangular parts, it is more probable that originally they were intended for leaves. In a few stripes the design is elongated, and in place of a single crossbar there are several, forming a figure that slightly resembles a poinsetta, which is the term occasionally applied to it by weavers.
The stripe shown in Plate [J], Fig. 2, which is found in Kazak and other nomadic rugs, is interesting as representing a vine of which the pendant flower is replaced by a T formed by latch-hooks.
In Figs. 3 and 4 of Plate [J], are patterns of stripes found in rugs of the Shirvan and Daghestan districts. As they are several centuries old, they may be derived from Armenian patterns, to which they show kinship. Both patterns are at times reversed as the result of copying.[31] A stripe also used in the same districts and probably of similar origin is seen in Plate [J], Fig. 5.
Differing from any of these because of their utilitarian origin, are the separate designs, which arranged in perpendicular rows, form the “churn” stripe of Plate [J], Fig. 6. Each of them represents crude machines for churning milk, which were formerly used by the nomadic tribes of Southern Caucasia and Armenia, who constructed them out of logs with a length of about five feet, and placed the sharpened base in the ground. Then hanging a goat’s skin filled with milk over each of the sides, and seating themselves in the middle, they turned first one then the other. As might be expected, these stripes belong entirely to nomadic rugs.
Figs. 7 and 8 of Plate [J], represent stripes sometimes seen in Shirvans. The latter is undoubtedly derived from the prayer patterns that are often used in these rugs.