The pile which distinguishes Oriental carpets from all others, is made by tying upon the warp, which has been previously stretched, wools of various colours, in such a way as to make a pattern. This is most dextrously done by the deft fingers of the Orientals, with great precision and skill, and the knots are called either Turkish or Persian according to the method of tying them.
For our purposes in rug-analysis it is not so necessary to know how these knots are tied as how they appear in the finished rug. On examination of the back of a rug we shall find that each thread of the warp is encircled by the knot-yarn, so that there appears a series of coloured stitches indicating the pattern which is worked out in knots on the surface of the rug.
Turning the rug so that we may see the ends of the knot-yarn which form the pile, we find that in some rugs the two ends reach the surface together between every other two warp threads, while in other rugs a sing-le end comes from between each two of the warp threads. The former of these is known as the Turkish or Ghiordes knot, the latter as the Persian or Sinneh knot.
The difference is at once evident upon investigation, and it may readily be seen that when the knots are so tied that one end of the yarn stands up between each thread of the warp there will be more knots to the square inch than when two threads of the warp are included in each tie of the knot. In all properly made rugs the knot is so securely tied that it is impossible to loosen or remove it by pulling the ends of the wool which form the nap or raised surface.
In this way the Oriental rug weaving differs from the attempts to copy the surface effect of the rug and according to orthodox methods, by drawing wools in and out of the warp without fastening them by knotting, so that the qualities of the fabric are not to be depended upon. It requires close examination to discover the knot itself in Oriental rugs.
On the back of rugs we find the encircling threads of wool, and on the surface the design is made of the ends of the yarn; so that we must separate these ends in our analysis and follow them to the warp, where we find the knots. In vast numbers of old rugs the pile has worn off so as to expose the knots themselves, which are so mosaic-like in character as to give name to a style.
Some collections consist wholly of such antiques, and it is absolutely impossible to reproduce their surfaces. However close the modern worker may cut the wools, and even burn away the ends with acids, the effect is unsatisfactory, and the attempt at deception is easily detected. Only age itself will produce the "mosaic style" so much coveted by connoisseurs.
Like beads upon a rosary, the knots seem to be strung, when an accidental ravelling gives us opportunity closely to examine the component parts of a rug, and a very good way of determining the claim of warp and woof to great age is to draw out a woof thread from any part of the rug and note how difficult it is to straighten it.
After days and weeks soaking in water, or even in prepared liquid, the kink still remains. It is true that some well-woven modern rugs may be thus tested, and the length of time taken to straighten the woof-thread may be almost as great as that needed by the antique ; but in a great number of specimens examined the result has been surprisingly convincing when other claims to antiquity have failed.
In fact, without seeing the rugs themselves, one becomes expert in discovering qualities and peculiarities of these woof threads which at first might strike one as being of the least importance of the three distinct parts of all pile carpets, the warp, woof, and knot. Fraud and a desire to lessen expense have led workers to introduce into the woof, which holds in place the knots after they have been tied, strands of cheaper materials than those used in the rest of the rug.
Threads of cotton are sometimes wound about by a thread or more of wool, and when the habits of weavers are learned these tricks are easily placed. Without woof threads there would be no weaving, and as both warp and knots frankly confess to the casual observer what they are, less attempt is made to introduce cheaper materials in them.
Heavy woof threads give weight and body to many beautiful fabrics, but it is when the woof threads are of good quality and extreme fineness that we find the most flexible results in the finished rugs. Prominence is sometimes given to fabrics by the introduction of metal. Upon a silk warp gold woof threads are woven, making a solid gold background for the knots of the pattern, which stands out in relief. Rarely beautiful is such unusual effect, and as temple-hanging or votive offering the creation is unsurpassed.
There will always be counterclaims made by enthusiasts for the greater ancestry of methods and designs. For many years Egyptologists argued with the lovers of Persian art for the supremacy of motifs of ornament, as well as of processes of manufacture.
The home of the knot carpet has been a matter of discussion, and, without lingering over any arguments for or against rival claims, we may safely assert that in Persia the fullest development of the art was reached, and from Persia the greatest inspiration was derived and carried East and West wherever the Oriental loom has been erected.
Within half a century the claims of China to priority in many art motifs and inventions have for the first time been severally considered, and much that was once ceded to Persia and India has been traced to China. Comparisons are now made between Persian and Chinese motifs of ornament that suggest similar former discussions between things Persian and things Egyptian.