the bow as it is drawn up, so as to be able to judge whether it bend evenly in both its limbs and show no sign of weakness in any particular point. The upper limb, as before stated, being the longest, should appear to bend a trifle the most, so that the whole may be symmetrical, when considered as bending from the real centre. It may next be tested, to ascertain whether it be a kicker; thus the string must be drawn up six inches or so and then loosed (of course without an arrow). If the bow have the fault of kicking ever so little, experience will easily detect it by the jolt in the hand. But on no account in this experiment should the string (without an arrow) be fully drawn and loosed. Care should be taken that the bow be sufficiently long for its strength. What has hitherto been said applies to all bows; but in self-bows attention must be paid to the straightness of the feathering of the wood. As a general rule, the lightest wood in a yew-bow will have the quickest cast, and the heaviest will make the most lasting implement. Between two bows of the same strength and length, the one being slight and the other bulky, there will be about the same difference as between a thoroughbred and a cart-horse. Therefore the preference should be given to bows that are light and slight for their strength. Light-coloured and dark yew make equally good bows, though most prefer the dark colour for choice. Fine and more open grain in yew are also equally good, but the finer is more scarce. If there be no bow suitable—i.e. none of the right weight—let the choice fall upon the best bow of greater power, and let it be reduced. Failing this, the purchaser may select an unfinished stave and have it made to his own pattern; but it is not easy to foretell how a stave will make up.

There remains one point about a bow, hitherto unnoticed, and this is its section, as to shape. This may vary, being broad and flat across its back, or the contrary—deep and pointed in the belly. Here again extremes should be avoided—the bow should in shape be neither too flat nor too deep. If it be an inch or so across the back just above the handle, it should also have about the same measurement through from back to belly. This much being granted, it is further declared that the back should be almost as flat and angular as possible, showing that it has been reduced as little as may be after the removal of the bark; but the belly should be rounded; and as the back should not be reduced in its depth towards the horns, and should not get too narrow across, it will follow that the chief reduction, to arrive at the proper curvature when the bow is drawn, must be in the belly, and therefore towards the horn. A well-shaped bow will in measurement become somewhat shallower from back to belly than it is across the back as it advances towards the horns.

Bows are broken from several causes: by means of neglected chrysals in the belly, or splinters in the back; by a jerking, uneven, or crooked style of drawing; by dwelling over-long on the point of aim after the arrow is fully drawn; by the breaking of the string; by damp, and oftentimes by carelessness; and even by thoughtlessness. Bows, moreover, may be broken on the steelyard in the weighing of them. A few years ago, when the Americans first took up archery very keenly, one of their novices wrote to a prominent English archer saying that he had broken nearly seventy bows in a couple of years, and asking the reason. He was told that he must either keep his bows in a damp place or the bows must be very bad ones, or else (to which view the writer inclined) he must be in the habit of stringing them the reverse way with the belly outwards. This would certainly have a fatal effect, but it is true that the Americans bought a number of very bad bows about that time from inferior makers in England. Whenever chrysals appear they must be carefully watched, and, as has already been said, if they become serious, a new belly must be added. This will not be a serious disfigurement, even to a self-yew bow. A splinter should be glued and lapped at once, but no one nowadays seems to care to have the covering patch painted as formerly, to represent as nearly as possible the colours of the different parts of the bow. Care should be taken not to stab the belly of the bow with the point of the arrow when nocking it; and the dents in the back of the bow made with the arrow as it is carelessly pulled out of the target should be avoided. A glove-button will often injure the back of the bow whilst it is being strung. As other ornaments—buttons, buckles, &c.—may also inflict disfigurements, it is better to avoid their presence as far as possible. Breakages from a bad style of drawing, or from dwelling too long on the aim, can only be avoided by adopting a better and more rational method. In order to avoid fracture through the breaking of strings, any string that shows signs of failure from too much wear or otherwise should be discarded; and strings that are too stiff, too hard, and too thin should be avoided. If a string break when the arrow is fully or almost drawn, there is but little hope for the bow; but if it break in the recoil after the arrow is shot, which fortunately is more frequently the case, the bow will seldom suffer. Yet if after the bow is strung the archer should observe that the string is no longer trustworthy, and decide to discard it, he should on no account cut it whilst the bow is braced, as the result of so doing will be an almost certain fracture. If the string be looped at both ends and the loop at either end be made too large, so that it slip off the nock in stringing, the bow may break, so that an archer who makes his own loops at the lower end of the string must be careful not to make them too loose. Breakage from damp is little to be feared in self-bows, except in localities where it is exceptionally moist, or, after long neglect, when damp has taken possession of the joint in the handle. In these cases single staves only are safe. Amongst backed bows there is much mortality from this cause. Commonly, it will be the lower limb that will fail, as that is most exposed to damp, arising either from the ground whilst shooting, or from the floor when put away. If the bow has been used in damp weather it should be carefully dried and rubbed with waxed flannel or cloth. A waterproof case, an 'Ascham' raised an inch or so above the floor in a dry room, and the bow hung up, not resting on its lower horn, are the best-known precautions. Half an inch of lapping, glued and varnished, above and below the joint of the horn is also a safe precaution against damp; also an occasional narrow lap in the course of the limb will assist to 'fast bind, fast find.' As regards the danger of carelessness, bows have been broken through attempts to string them the wrong way, or by using them upside down; and thoughtlessness will lead the inexperienced to attempt to bring a bow that follows the string upright, to its infinite peril. In such cases the verdict of 'Serve him right' should be brought against the offender if he be the owner. In weighing a bow on the steelyard care must be taken to see that the peg indicating the length to be drawn be at the right point; otherwise a lady's bow, for instance, may be destroyed in the mistaken attempt to pull it up twenty-eight inches, or three inches too much.

It has already been stated that a belly much injured by chrysals may be replaced by a new belly; any incurable failure of the back may also be cured by its renewal. A weak bow or limb may also be strengthened by these means. Also, if either limb be broken or irretrievably damaged, and the remaining one be sound, and worth the expense, another limb may be successfully grafted on to the old one. If possible, let this be an old limb also, as the combination of new and old wood is not always satisfactory; the former (though well seasoned,