384 Memoirs Bcniicc P. Bishop Museum 



Edge Partington (20, p. 46) fignres a closely woven basket of coconut 

 leaves of still a different type. It is oval in outline, with a flat bottom, vertical 

 sides and a flat recurved top, making the opening considerably smaller in dia- 

 meter than the body of the basket. Xo baskets of this sort are now in use and 

 the form seems to have been forgotten. 



In the collection of the Peabody INIuseum at Cambridge there are two 

 rather curious sacks from the Marquesas. 



The material is split pandanus similar to that employed in mat weaving. The sacks are 

 ahnost identical in size and construction and are rectangular, \^y2 inches long and 11 inches 

 deep. The form is that of an envelope with a triangular flap. The body of the sack is of 

 double thickness, the outside layer being :riade of strips 14 i"ch wide except for the last 

 T, inches below the top. The sack above this point, the flap, and the inner layer, are woven 

 frot7i much finer strips produced by splitting each of the 14 in<^h strips in three. The tech- 

 nique is a simple checkerwork. A loop of twisted coconut fiber is attached to the center 

 of the lip of the sack and passes up through a hole in the flap, protruding for 2 to 3 inches 

 when the sack is closed. There is no information as to the use of these objects. 



Jardin (},^, p. 36) says that the natives wove coarse baskets from the 

 flexible shoots of the tutu (Ceanothus asiatious), but gives no data to their form 

 or the technique employed. 



FAXS 



Marquesan fans were of two sorts, ordinary fans, which are still manu- 

 factured and used, and chiefs' fans, which were carried by persons of importance 

 and were intended for show rather than use. The chiefs' fans are now obsolete, 

 and no examples have been preserved in the grou]i. Thev are fairh- numerous 

 in collections, however, and the type may be accttratelv described. 



The coconut fans have a broad straight outer end and sides which arc 

 nearly parallel at the tip and curve inward smoothly at the base. All the speci- 

 mens seen were more or less asymmetrical (PI. Lxvi, F), but this did not appear 



to be intentional. They are made as follows: 



Young coconut leaflets are folded so that the midrib lies along one edge of the woven 

 strip. The weaving is begun at the outer end. a flat band, twilled over two and under two 

 and as long as the width of the proposed fan, being woven first. The twill in this band is 

 longitudinal, and the whole is rarely more than seven elements wide. The direction of the 

 twill is then changed to vertical, and the fan woven down to the upper end of the handle. To 

 form the handle the two leaflets which are central at that point are crossed and carried 

 down vertically, each succeeding pair being similarly detached from the weaving, crossed and 

 braided in with those already detached until all have been included. The ends of these leaf- 

 lets, which extend some distance below the fan, are then packed closely together, wrapped 

 with bark string, and cut off squarely at the tips, the whole forming the handle. 



The following description of chiefs' fans (PI. lxvii, A. B) is based upon 

 an examination of specimens in the possession of I\I, George La Garde of 

 Papeete, and in various American museums. The weaving appears to have been 

 uniform ; the technique being as follows : 



[124] 



