Part I

the island of Nunivak (Fig. 6, No. 15651, collected by W. H. Dall. This is an unusually large and stout bow, with 37 strands in 310 REPORT OF NATIONAL MUSEUM, 1884. the cable), and one from near Cape Eomanzoflf (Fig. 7, Iso. 36034, col- lected by E. W. i^'elson), the strands are twisted from the middle (the two ends of the bow in opposite directions) by introducing a toggle, between the strands, and the twist is secured by passing the "stop^ through the cable. -' This type extends from the island of Kadiak toj^orton Sound. The second form of the type appears to be less common than the first, though occurring alongside of the latter. It appears not to be used on the island of Kunivak or south of the Kuskoquim Eiver. II. The Arctic Type. This is a much shorter bow than the above (from 43 to 52 inches long), narrow in proportion, and of a much more graceful shape (Fig. 8, Iso. 1972, from the Mackenzie region, collected by Eoss). In section it is nearly elliptical, flatter on the back than on the belly, with the handle slightly narrowed and thickened. The greatest breadth is usually about IJ inches, and the thickness at the handle about f inch. The jends are often bent up as in the second form of the southern type^ and when this is done the back is usually reinforced with a short rounded strap of wood or antler in the bend. One bow (Fig. 9, a^o. 89245, froin Point Barrow, collected by our expedition) has these ends made of separate pieces mortised on. Only one bow of this type in the collection has an extra rib, which is of antler and very small and short, but the back is frequently covered, with.jtiipa of sealskin, put on length- wise. The backing is always of braided sinew, and of a very complicated and perfect pattern, usually very thoroughly incorporated with the bow by means of hitches and a very complete seizing of many turns running nearly, the whole length of the bow and serving to equalize the distribution of the strain and thus prevent cracking. The backing is one continuous piece of cord, except in one case, where the seizing is separate, and begins, as usual, with an eye, which is slipped round the upper nock. The strands vary in number from 30 to 45 on a man's bow (22-28 on a boy's) of which 10-26 extend only from bend to bend on a bow of the Tatar shape, or between the corresponding points on a straight bow, and are then made fast by two or three half-hitches each, or, as at Point Barrow, Wainwright's Inlet, and Point Hope, by complicated lashings made up of series of half-hitches, ^ften alternately in opposite directions, the last hitch or two held down Jby extra round turns, and sometimes as many as a dozen hitches in a series. Fig. 10 is this section of the same large bow, No. 89245, from Point Barrow, figured above, and Fig. 11, the same part of No. 72771, from Wainwright's Inlet, also collected by our expedition. A detailed description of the lashings of these bows, two of the most complicated in the collection, will make these figures plain. The first ; ESKIMO BOWS. 311 long strand on reaching the bend is hitched round the bow seven times at intervals of about f to 1 inch. These "under-hitches," as they may be called, occur always on bows of this tyx)e, sometimes made by the first and sometimes by the last long strand, and serve to mark off the position of the hitches of the short strands and give them a point (Tappui. The first two of these are " two half-hitches," or, " clove- hitches," as they are called at sea, the other five peculiar hitches (Fig. 12) not used by sailors. The hitch is well known and much used in the artillery and ordnance service, and is there called a "clove-hitch." As using this name would not distinguish the hitch from the common " clove-hitch " of seamen, I venture to suggest for it the name of " sol- dier's hitch." It is made by taking two round turns round the object to be fastened to and bringing the end over the standing part and under the two turns. If the turns are taken to the left, it makes Fig. 12 if to the right, Fig. 13. These hitches, especially the left-handed one, are much used by the Eskimos not only on bows, but in putting on seizings upon spears, &c., where a white seaman would use a " marling-hitch." The advantage of this form of hitch seems to be that the second round turn keeps it from slipping if the end gets loose. To return to No. 89245 : After making " under-hitches" at both bends, long strands are laid on till there are ten in all. The eleventh, on reach- ing the bend, makes two "soldier's hitches" at 1, and going to the other nock is similarly hitched at the other bend, and then passes back- wards and forwards between the bends, hitched each time nearer the middle of the bow. The hitch at 2 is made thus : Two round turns to the left, the end passed under both turns, and then two more round turns, with the end passed over the second turn, under the first and third, over the standing-part and third turn, making a double " sol- dier's hitch." At 3 are two simple half-hitches, and one made with two round turns, followed by two round turns with the end passed under both. At 4 is a similar lashing with eight simple hitches ; at 5, nine ; at 6, four ; and at 7, two. In Iso. 72771 there are five " under hitches," all " soldier's hitches," made by the first long strand. The lashing at 1 is made by hitching alternately to right and left five times. (Such hitching is caUed " kack- ling" by seamen.) At 2 it is "kackled" nine times, at 3 nine times, and at 4 nine times again, ending with a half-hitch at 5. It will easily be seen, as was suggested to me by Professor Mason, that the strain of bending the bow, while tending to stretc^&~ahd tighten each longitudinal strand, at the same time tightens each individual turn of these lashings, so that the greater the strain on the bow the tighter do they grip the fibers of the wood and hold them together.. These hitches usually occupy 4 to 6 inches of the bow, and as a rule are put on as above, so that the shortest strands come at the top of the 312 REPORT OF NATIONAL MUSEUM, 1884. backing, though they are reversed on one bow from the Mackenzie re- gion (Fig. 14 is this section of No. 1970, collected by Eoss), so that the longer of the strands are stretched across the bends, which adds some- what to the tension of the bow, but makes a less neat and compact lashing than the common arrangement. This arrangement of the shgrf ^s strands brings the greatest strength across the middle of the bow, where it is most needed. All the strands between the hitches are divided into two equal ijar- <y eels and twisted from the middle into two cables, thus greatly increas- ing the tension to be overcome in drawing the string. These two cables are fastened together by a sort of " figure-of-8 " knot, passing through and around them, and are stopped firmly to the handle, after which the whole is securely seized down with the end of the backing. This seiz- ing is less complete in bows from the region of the Mackenzie. In one case, after completing the seizing the end goes on to lay on a few strands more, for a third cable, outside of and between the other two, which is also twisted. (No. 89245, Figs. 9 and 10. End of cable cut off at a.) The ends of the long strands, between the nocks and the hitches of the short strands, are sometimes wound with separate pieces. Bows of this pattern, differing only in details of the backing, are used at the Mackenzie River, at Point Barrow, Wainwright's Inlet, Point Hope, and the Diomede Islands in Bering Strait, and probably at" intermediate points along the shores of the Arctic Ocean. As was said above, there are no bows in the collection from Kotzebue Sound or the Kaviak Peninsula, but from several points in the region in question, namely, from Kotzebue Sound, Hotham Inlet, Sledge Island, and Cape Nome, have been obtained many of the ingenious little tools for twisting the cables, and always in pairs, indicating that a two-cable ^ bow of the Arctic type is the prevailing if not the only weapon of the tlnd used in these localities. The line of demarcation between this type and the preceding is not sharply drawn, although there are no bows of the pattern which is ex- clusively used as far north as Cape Romanzoff, in the collection from north of Bering Strait. From the Yukon delta we have one bow (Fig. 15, No. 33867, collected by E. W. Nelson), which in proportional narrowness and thickness ap- proaches the Arctic model, as it does in its complete seizing, though it has a strong extra rib, and the general pa:tern of the backing is purely southern. From the same region is another