BPSSELL] ARTIFACTS 151 



lining a rcctan<jular space the exact size of the j>rojpcted fal)ric. A 

 deep layer of clean white sand from the river bed was spread between 

 the stakes to prevent the under surface of the fabric from becoming 

 soiled. The end or yarn ])eams were of saj!;uaro ribs of suitable size, 

 held in place across the end stakes by cords which were stretched taut 

 at the sides. The beams were about 6 inches from the around, tlms 

 perniittin<^ the warp to pass freely around them as it was wound over 

 one and under the other in a continuous thread. A heavy double 

 binding thread, usually dyed buff, was passed through the loops at the 

 ends of the war]) and was given a half turn as each loop was caught up. 

 The yarn beams were then removed, leaving a lease rod of arrowwood 

 in place of one of them. The binding thread was next bound to the 

 yarn beams by a heavy thread wound in a spiral from end to end. 

 The warp could then i)e stretched in place !))• again putting the beams 

 outside the stakes and pulling the side cords taut. The heald rod was 

 also of arrowwood put in i)lace by passing a loop from a thread that 

 had been slipped through the open shed from the right imder each 

 lower warp thread and pushing the rod through the loop from the left. 



The weaver sat upon his haunches on the ground or on the cloth 

 when it was finished too far for him to reach frt)m the end. He low- 

 ered the lease rod beyond the heddle and gathered the upper threads 

 in front of the heddle on a slender sharpened rod, which enabled him 

 to lift them higli enough to pass the shuttle through. The shuttle was 

 an arrowwood stick to one end of which the weft thread was tied and 

 then passed to the other in a slightly sjiiral direction; there it was 

 wound twice around and then j)asseil back; thus it was wound from 

 end to end of the shuttle until the latter carried many yards of thread. 

 After the shuttle was j)assed through the shed the thread was struck 

 home with a flat batten of mesquite wood. If the war]) threads were 

 irregular, thej' were adjusted with a short peg which took the place of 

 the comb used by the other tribes that weave in the Southwest." Two 

 heavy selvage threads lay at the side of the warp, and as the shuttle 

 was passed through the shed it was brought between them and they 

 were given a half turn to engage the woof thread before it ])assed back 

 in the other shed which was ojjened by raising the heddle. 



The width of the cloth was not well maintained, as there was a strong 

 tendency for it to become narrower, but by the aid of a stretcher or 

 tem])le this was ])artially overcome. The temjile had two longitudinal 

 grooves separated by a distance equal to the w idth of the clotii. In 

 eacli groove was ])laced a s(>ction of willow or arrowwood stem an inch 

 in length, bound with a heavy thread around the stretcher. The outer 

 end of the short stick was sharpened so that it might be ])ushed 



a Length of the specimen collected. U cm. U is shown in fig. 72 at the margin of the finished cloth, 

 where it was pushe<i under u few threads to hold it in position for photographing. 



