Again, the substitution of other materials for the bark layer as a vehicle for the ornamental lacquer does not, judging from this specimen, appear to be a success.

The Indian bows vary to a considerable extent in form and in external appearance, but, so far as I have been able to see

from superficial examination, the greater number do not differ materially in structure from the specimen described, which may be taken as fairly typical of the class. Many approach more nearly to the Persian type, and bark is often present beneath the lacquer.

I have not been able to examine a number of Turkish bows, but I gather that they are for the most part only slight modifications of the type of which the Persian and Indian bows are varieties.[20] A bow in the British Museum, described as Turkish, is small and very beautifully finished. Its length is 3 ft. 8½ in., and greatest width 1 1/5 inches. The grip is covered with bark, and bulges towards the back only; the horn of the belly is exposed and polished, thus reminding one of the "Tatar" type; it is in two pieces, separated at the centre of the grip by a thin ivory plate. The sinew backing is covered with thin black leather, upon which designs are picked out in gold. The ridges are strongly marked, and the "ears" of plain wood and very short, partly covered on the back with birch bark. The nocks are lined with leather. In transverse section the arms are plano-convex. It is very powerful for its size; the reflex curve is very regular and increased gradually towards the "ears," resembling the curve of most Persian bows. Of the internal structure I am unable to speak.

I have confined myself in the above remarks to the class of weapons which goes by the name of the "composite bow," that is, bows which have a reinforcement of sinews on the back, and which in many cases exhibit further a composite structure, in the presence of a variety of materials. There are, however, a few forms which, although they must be excluded from this class, nevertheless show a relationship to the composite type, and give evidence that they have been derived from it.

Many plain wood bows from the Oregon Indians have a strong reflex curve when unstrung, though this is not due to the presence of sinews on the back, the curve being carved to shape in the wood itself. They are very flat, short, and springy, and in general character suggest relationship to bows of composite nature. Besides these, most of the bows of the Clapet tribe show a similar relationship. These again are plain or "self" bows, but in their strongly plano-convex or concavo-convex section seem to point to the aforesaid origin. They are moreover characterized by having broad grooves along the back, which may be considered as possibly imitating a former sinew backing, or even as being channels along which a sinew cord used formerly to lie, though now disused and merely retained from force of habit.