In Persia there is also an instrument strongly resembling the Omerti and Kemangeh in outline, called the Sitara (Fig. 9). Then there is a primitive bowed instrument with three strings, known to the peasants of Russia as the Goudok, which is no doubt an immediate descendant of the three-stringed Rebab, and, more remotely, of the Ravanastron. Abyssinia too, has its bowed instruments. In fact, the use of the bow is universal in the "glorious Orient," from whence nearly all products of western civilization are derived. In almost all cases great antiquity is ascribed to these instruments. The very name "Kemangeh-a-gouz," ancient in itself, can be roughly translated "ancient-fiddle," thus showing that the Persians [the name is Persian and bears out the Arab records that it came to them from Persia] considered it then a relic of the past, and that it was a survival of some still older instrument inherited, most likely from India. There can be little doubt that Fetis was right in assuming this to have been the Omerti, for, barring the long "tail-pin," the structure of both is almost identical. [Illustration: FIG. 9.] The bows of all these instruments bear a strong resemblance to each other, as is only to be expected where all are of the simplest description. In the majority of cases the bow is merely a length of cane with a bunch of horse-hair tied at each end in such a manner as to pull the cane into a more or less pronounced curve. Those of the Goudok and Sarinda (Fig. 10) are short, approach nearly to a semi-circle, and are quite rigid. [Illustration: FIG. 10.] Those of the Ravanastron, Omerti, etc., are longer, and being more slender, have a certain amount of flexibility, but it does not appear that this latter qualification is sought for or considered indispensable. On the other hand, the now nearly obsolete Kokiu of Japan had a bow of about forty-five inches in length that was extremely elastic. It was made in sections after the manner of a fishing-rod, and the hair was tightened by the finger of the player, as in some of the early viol bows of Europe. The method of hairing in most cases amounts to the simplest way of tying the hair on to the stick. Sometimes the hair is passed through a slit and held in place by a knot. In other specimens it is attached to a leather thong, and occasionally it is plugged into the open end of a piece of bamboo (Fig. 11). [Illustration: FIG. 11.] The bows of the Saw-tai (Fig. 12), Uh-Ch'in, Koka and a few others show a distinct advance in point of curve and adjustment of hair, and strongly resemble the bow of the quaint Swedish Nyckelharpa in present use (Fig. 13). [Illustration: FIG. 12.] [Illustration: FIG. 13.]