Feathering.—The feathering is made by splitting a strong feather and spacing three of the parts at the base of the arrow. A turn of sinew is first made to hold them in place, then they are bound on and trimmed to a gauge. Some tribes also fasten the feathers with glue; others bow or spring the feathers to aid the flight. The Point Barrow Eskimo fasten the feathers by forcing the extremities of the feather shaft into the soft wood with an ivory punch. In African arrows the wrapping thread is carried through the barbs of the feathers without disarranging them. Most arrows have triple feathering. The Eastern Indians can only be conjectured to have followed this rule, since there are no specimens of this art in existence and it is doubtful if records are minute enough to note this point. Some of the Cherokee arrows collected by Mr. James Mooney, however, have two feathers and are rifled. The Dog Rib and Iroquois Indians also used only two feathers. The Northwest Coast arrows have two entire stiff feathers lying flat against the sides of the shaft, seemingly devised to rotate the arrow. This style begins on the Columbia river and, with a few exceptions, follows the Eskimo coast fringe to Labrador. A notable exception occurs on the Yukon river and delta, where superior arrows are made. The Pima of Arizona are the only Indians south of Oregon in the western United States who employ a two-feathered arrow. The double feather reappears in British Guiana, on the Amazons, and in Terra del Fuego. In the first two localities mentioned rifling is also rather common. African arrows are the only ones in which I have observed multiple feathering, running from triple to octuple or higher.

In general, rifling the feathering is unusual. In many cases it is, perhaps, an adventitious result, though undoubtedly designed in the Pima arrow and in one specimen from the McCloud River Indians. If the Pima and Mohave arrows are all rifled, as indicated by the Museum specimens, the double feathering is explained, since rifling by means of three feathers would render the aim bad by jostling the arrow when the feathering crossed the bow. With two feathers a clearer space is left. In the Pima arrow the rifling is designed, as the feather from before is but slightly curved for some distance, then is lashed with sinew, bent more sharply, and lashed again, etc., until the turn is quite marked near the nock. Nearly the same plan is pursued in the McCloud River arrow. The rifling curve is usually small, being rarely over 90° in 4-5 inches.

It seems not a little remarkable that rifling, which is supposed to be a modern invention, should be discovered and practiced by savages; but it is undoubtedly true; and it is possible that the "beveled" arrow-heads of camp sites were designed to effect the same rotary motion.

It is the usage of some Indian tribes to leave a trail of long barbs on the feather near the nock. Very often the Sioux allow the downy plumules at the junction of the shaft and quill of the feather to remain, giving a finish and probably a tribal mark. The ornamentation of the arrow also affords tribal marks, notably among the Indians of the United States.