434 Memoirs Bcniicc P. Bisliof' Miiscutn 



head in addition to the large taai'alia." 'I'his probably refers to ornaments of the 

 same general tvpe as that shown on Plate Lxxx, A. 3, which were made of un- 

 usually long tail feathers of the tropic bird attached to a smooth stick i)v wrap- 

 pings of fine sennit. The principle of attachment appears to be mnch the same as 

 that used in the large taavalia except that the plumes entirely surround the stick 

 instead of running only along one side of the band. The lower end of the care- 

 fully smoothed stick projects for some distance, suggesting that the ornament was 

 thrust into a coil of hair. 



Perhaps the most beautiful of the Marquesan headdresses was the packua, 

 shown on Plate lxxx, B. Only two of these ornaments were seen, but these agree 

 so closely in all details of construction and decoration as to suggest that the 

 type was a fixed one. It is rather interesting that one of the specimens contains 

 a few peacock feathers, pro\-ing it to be i^ost-European. 



The Ixxly of the packiia is made up of three strips of featherwork. Each strip is broad- 

 est in the middle and tapers to a point at either end. A piece of large bamboo, flattened and 

 scraped to paper thinness serves as a base for the featherwork. Over this are laid one or 

 more layers of fine white tapa, which fit the bamboo so closely as to suggest that they were 

 put on while the tapa was wet. The feathers, all of which are small, are glued to this tapa 

 with breadfruit gum. To give additional securit}' the whole strip is wrapped with single 

 strands of some stitt white fiber — probably banana fiber. In a perfect specimen these wrai)pings 

 are concealed by the feathers, but in the one figured the wrappings, tapa pad, and bamboo 

 base are all visible. The feathers are laid on in irregular overlapping rows, the work progress- 

 ing from both ends and meeting in the middle. 



The tapa pad varies in thickness in the different strips composing the ornament, being 

 heaviest in the top and bottom strips, less heavy on the ends of the central strip, and thinnest 

 at the center of the central strip. 



The three strips are pasted side by side upon a single large piece of coconut leaf 

 sheath, whose natural stififness serves to hold the ornament in shape. Along either edge, out- 

 side the strips, run three flat bands, or ribbons, plaited from strands of black dyed fau bark 

 and white tapa. At either end of the ornament, between these bands, is a small triangular 

 piece of white tapa which serves to cover the tips of the strips of feather work. Long tying 

 strips, braided from tajia and black fau bark, are attached to either end of the ornament. In 

 the specimen figured two tapa strings are fastened to the inside near the top center, suggest- 

 ing that there was originally an additional ornament which projected above the packiia. 



Green, yellow and red feathers were used. The first two colors were obtained from a 

 species of dove, the kukii. The source of the red feathers is doubtful. The usual native 

 story is that they were obtained from the manu kua. a red bird long extinct, wdiich lived on 

 the plateaus of Xuku Hiva and Hiva Oa. In an account obtained in Pua Ma'u, Hiva Oa, they 

 are said to have been brought from a far oflf island called Aotona, which Handy identifies as 

 Rarotonga. A species of kuku which has a red cap of the same shade as the feathers in the 

 packua occurs in Hiva Oa, and it seems probable that at least some of the feathers used were 

 obtained from these birds. ( It should be noted that Dordillon defines packua as "head orna- 

 ment of kuku feathers.") 



The red, green and yellow feathers on each strip were arranged in vertical bands, the 

 center of the middle strip being covered by a large block of red feathers. Two small designs 

 in red, one of which appears in the illustration, were placed on either side of the centers of 

 the outer strips. 



The packua is said to have been the ancient headdress of Xuku Hiva and also of the 

 vallev of Pua Ala'u in Hiva (.)a. It is not mentioned by any of the early writers, however, and 



[174] 



