The problem of the actual length of the composite bow of the Arabs is a baffling one. The span is given by most authorities as nine inches but, by actual measurement of hands, we have found that the average man can stretch his thumb and little finger not much over eight inches. If we accept the nine-inch span, the length of five one and a half spans would be sixty-seven and a half inches, and of five one and two thirds spans, seventy-five inches. The eight-inch span would give sixty-two and a half and seventy inches.

While such measurements would be quite acceptable for wooden bows, they certainly seem too long for the kind of Oriental composite bows of which we have seen examples, or have read about. And let us state at once that all our efforts to find a genuine Arab bow have so far been unsuccessful. We must form our opinions on comparisons, calculations, and a few questionable pictures.

The statement on page 77 of this volume, that "the length of the bow is three cubits and one finger," or sixty-five inches, agrees perfectly with the five-span measurements given above; yet how could one reconcile such a length with the statement on page 87 that "the arrow should measure exactly the length of the strung bow"? Evidently different sorts of bows were cited, the short ones being quite likely from Persian sources, or authors, and the long ones from Arabic.

The longest composite bow that we know of is the Chinese, which is seventy-four inches from end to end. However, such very long bows lose all the advantages that should be gained by composite structure and have a cast that is very poor in proportion to their weight: only a fraction of what the short Turkish and Persian bows can do.

As good specimens of short composite bows, we personally measured three beautiful ones from India that were exhibited in the Centennial Exhibition at Philadelphia in 1876. Their details are: Bow 1. Grip: 4 inches; arm: 10 inches; siyah: 9 inches. Total: 42 inches. Bows 2 and 3. Grip: 3 1/2; arm: 9 1/2; siyah: 6 1/2. Total 35 1/2 inches.

The best Turkish bow of Sir Ralph Payne-Gallwey, author of A Treatise on Turkish and other Oriental Bows (1906), measured forty-four inches when unstrung and the bowstring was thirty-three inches in length, which would suggest a total strung length of perhaps a yard. A modern Sind bow in our collection, which is a masterpiece of bowyery in the ancient tradition, measures fifty-six inches.

In the Metropolitan Museum of Art are many Persian miniatures which show the composite bow very accurately: strung, and at full draw, but not, so far as I know, unstrung. Most of the bows that are strung, but at rest, are seen projecting from their quivers, which are hung on the right side of the archers exactly as described on page 155, but in a painting which is dated as prior to 1575 and is labeled: "Assad ibn Kariba attacks the army of Iraj suddenly by night," a very carefully limned archer holds a beautiful strung bow at rest. By measuring along its curvature with a fine flexible steel tape we found that its length was equal to the distance from the ground to the archer's nipple which, for a man of sixty-eight inches, would mean about fifty inches. It seemed to us that the assumption of accuracy in proportion was warranted.

Inasmuch as the Arab composite bow occupied an intermediate position between the Hijazite Arab and Persian bows, it was doubtless shorter than the former and longer than the latter. The direct statement that it measured sixty-five inches cannot be dismissed, especially in view of the great length of the arrows which is described in Section XXXVII.

Some collateral evidence is afforded by a Frenchman named Belon, writing in 1553—which is not so far from the date of the present manuscript—who said:
"The bows and quivers that the Arabs carry are different from the others of Turkey. The bows of the Arabs resemble more the Grecian bows than the Turkish bows, for the Turks of Asia carry a little bow well braced up, strongly curved, and very stiff. But the bows of the Cretans are of two sorts, of which those made in Sphagia [or Sphacteria, a small island off the southwest coast of Greece] with the horns of the ibex, and those made in Candia [Crete] with the horns of buffaloes, are larger than the Turkish; and since they are larger than the Turkish, so must they have long and thick arrows quite as much as the bows of the Arabs; for the Arabs have their bows big because it is necessary for them to use large arrows, contrary to the Turks, who have theirs small." (Archers d'autre Fois, Stein [Paris, 1925].)