A more mechanical way is to give the twist by a spindle bearing a whorl.
Our first problem is to distinguish the different methods of spinning and state their respective distributions. In this we must proceed with hesitancy because of inadequate data, but since very little native spinning survives there is no ground for expecting important additions to our field observations. The subject is, therefore, ready for such comparative study as can be made. We have neither the time nor the special knowledge to do this now, but will discuss the most obvious points.
One of the most direct approaches is the distribution of the spindle whorl. Its known occurrence in North America is in the highlands from Panama to the Colorado River. Then with a break in continuity we find it in British Columbia and on the adjoining coast. The only other place where there is even a suspicion of its use is the lower Mississippi. This conjecture is based upon the bare mention of an improvised affair, a wad of clay upon a stick, by an early writer.[1]
Archæological data are on the whole consistent with the foregoing facts, from which their general finality may be assumed. Pre-Columbian sites yielding undoubted spindle whorls must be our safest criteria, because we lack definite knowledge as to the exact state of spinning before white contact, and it is conceivable that the use of the European whorl could have been introduced quickly, as we have already noted in case of the horse. Peru presents a puzzling case, for notwithstanding the high development of the art, the early historical data indicate the absence of a true whorl,[2] and objects of this nature are seldom found in excavating. However, in the northern part of the Andean region, pottery objects resembling whorls are frequent. In other parts of South America they are rare, but the modern natives use them. If we try to correlate the distribution of the whorl with that for fibers, we note that it is wanting in the distinctively bast and sinew areas. Where cotton, wool or both together are spun, we find the whorl, unless Peru should prove to be an exception.
Outside of the whorl area we have defined, bast and sinew thread are given the final twist by rolling under the palm of
