As the scale is ascended the tendency to conceal the structural details beneath an external coat, and thus to give an homogeneous appearance to the bow, becomes increased. We observe it in its infancy in the Siberian bows with their plain or very slightly ornamented bark covering, lying over the sinew backing; and higher in the scale this coating, which at first doubtless served a purely useful purpose, as a protection from the effects of weather, becomes more and more a vehicle for the embellishment of ornamental art, at the same time increasing in its extent, till the maximum is reached in bows of the Persian type, in which usually the elaborate structure is entirely concealed by a coat of lacquer, upon which frequently great artistic skill is lavished in floral designs and scroll work picked out in gold. All composite bows appear to require soaking in water to produce their maximum effect, and possibly this bark coat, besides protecting the sinews from injury, was intended primarily to prevent rapid change in the condition of the bow, and especially the sinew and glue, from changes in the temperature, and to protect them from the sun's rays. Secondarily, it was found to be a convenient ground upon which to lay the varnish and paint which give the finishing touches. I do not know the composition of the lacquer used, but it must doubtless be of a very special nature not to crack all over when the bow is bent.

The specimens figured are, as appears to be usually the case with the Persian bows, entirely covered with the lacquer coat, except at the edges of the arms, where the side strips of horn appear on the surface, as in most specimens where they occur at all.

A section (Fig. 10) taken transversely across the centre of one of the arms, at once exhibits a marked difference from the corresponding section in the Tatar bow (Fig. 4). It is seen that the centre (a, a) is composed of a light-coloured wood in two pieces, unequal in width, and the surface of this is much scored with rough grooves, to give a firm hold to the glue and sinews. The belly is composed of a number of narrow strips of horn (b, b) instead of a single piece. These are joined to the wood and to each other with glue, which is seen filling up the interstices as an hyaline substance (h). Over the horn strips is a very thin layer of transversely disposed sinews mixed with glue, extending from side to side, and apparently to assist in keeping together the numerous strips. This does not occur in the "Tatar " bow. The back is covered with a thick layer (c) of longitudinal sinews, slightly mixed with glue, the layer being well coated on the outside with glue, the surface of which is smoothed and polished. The sinew layer appears to be single and not in two strata, as in the "Tatar " bow. Overlying both belly and back is a layer of the finest inner bark of the birch, very delicate, and applied in rhomboidal pieces, as before described (there is no layer of coarser bark), and immediately upon this lies the external coat of lacquer. At the edges the strips of horn (d, d) are exposed and break the seemingly spiral winding of the bark, which is only apparent, as the edges of the pieces on the back and belly do not correspond.

Fig. 11, Plate VI, is taken from a dissection of the belly side of this part, shewing the successive strata—the horn strips (b, b); the external side strips (d, d); the transverse sinews (k); the bark layer (e, e), shewing portions of two pieces; the external lacquer (l), which replaces the bark coat (f), of the "Tatar " bow. A dissection of the back is shown in Fig. 12, where c represents the sinew reinforcement, and c2 the external surface of this, coated with smooth polished glue.

Fig. 13 shews a transverse section through the middle of the ridge at the commencement of the terminal"ear," corresponding to Fig. 6. The number of horn strips is smaller than at the centre of the arm, shewing that these do not all run the whole length of the arms; the horn ends abruptly at the commencement of the "ears," about three inches beyond the point at which this section is taken. It is also seen that in the Persian bow the wood base enters less into the formation of the ridge than is the case in the "Tatar" form; the ridge is here almost entirely moulded up from the sinew mass.