Linton — Tlic Marquesas Islands 379 



Finished three plait cord is stored in long cylindrical rolls wound to give 

 an ornamental effect. Many strand cord on the other hand is said to have been 

 commonly wound in thin flat rolls like European tape. (See PL Lxv, B.) 



The only dyed coconut fiber cord collected was of the three strand 

 variet}^ but it is probable that the more intricate plaits were also occas- 

 ionally colored. All the specimens seen were reddish brown, tan or black, 

 but red is mentioned in native stories and Stewart, a reliable authority, speaks of 

 yellow and white sennit used in house lashings. The reddish brown and tan 

 sennit appears to be natural. The Tongan method of deepening the color by 

 baking, if originally known to the natives, has been forgotten. The lightest 

 shade of tan sennit was made from fiber of the leaf sheath. Three methods of 

 coloring the black sennit were described. 



The bark of the iiiai'i tree was beaten up with stones and the juice squeezed into large 

 wooden bowls. The finished sennit was then steeped in the juice for several days, until it be- 

 came black. It was then removed and buried for three days in the mud of a tare patch, 

 washed, and dried. A second method consisted of burying it in the mud for a longer period 

 without the preliminary dyeing, while a third method consisted of rubbing the finished cord 

 with a mixture of ama nut soot and water. 



It seems probable that when mention is made to red sennit the darker 

 shades of red brown are referred to, but at least some red sennit was made by 

 painting the fiber, apparently after plaiting, with the seeds of the arnotto {Bixa 

 orellana). Yellow sennit was colored with saffron, the cord probably being made 

 from the light colored fiber of the leaf sheath. White sennit must , from the 

 nature of the fiber, have been incrusted or painted rather than dyed, and it is 

 cjuite possible that it was produced by rubbing the finished cord with a paste 

 made from shell lime or from a fine white earth, extensive deposits of which 

 occur in Nuku Hiva. This supposition is strengthened by the fact that fans were 

 whitened in this way. 



Although most Marquesan cordage was made from either fau bark or coco- 

 nut fiber a number of other fiber plants were known and used. The most im- 

 portant of these seems to have been the pineapple. Pineapple fiber is still used 

 for thread and for very small fish lines, and is prepared as follows : 



The fresh green leaves are taken one by one, laid across a banana log, and 

 scraped longitudinally with any sharp-edged instrument. The bowl of an ordinary metal spoon 

 was seen used for this purpose. Continued scraping removes all the flesh of the leaf, leav- 

 ing a few soft fibers which are then dried in the sun. The women convert these fibers into 

 thread by rolling them upon the bare thigh with the palm of the hand. To make fish lines 

 either two or three strands of this thread are twisted together, the finished line being passed 

 rapidly through a flame to singe off the lint. 



Pjanana fiber is said to have formerly been used for certain kinds of cord, 

 but this material has not been employed for many 3'ears. To prepare the fiber, 



[119] 



