Linton — The Marquesas Islands 443 



"'•••'■''•'• STRING RECORDS 



' The use of string records sterns to have been more highly developed in the 

 Marquesas than in any other part of Polynesia. Such records served principally 

 to aid the memory in the recitations of chants and geneologies, but Stewart (59, 

 p. 335) metitions a time record consisting of a string of tapa in which a knot was 

 tied at the. return of every full moon. One complete string record, and part of a 

 second were collected by the Bayard Dominick Expedition. 



I , Tli^e complete record, shown on Plate lxxxiv, E, obtained in Pua Ma'u, Hiva Oa, 

 has a cyHndrical body and is made from rather small three-strand sennit applied in a wrap 

 weaV'ei '' 'The warp strands, which run longitudinally, are placed on a slant so that each strand 

 encirdjes , the ,c>'linder once between the upper and lower ends. The upper end tapers to a 

 point and terminates in a loop. At the juncture of this tapering portion with the body of the 

 cylinder ' there are five small loops of sennit, placed at equal intervals on the circumference, 

 to each of which is attached a small piece of white tapa. The lower end of the cylinder is 

 Open,: and .its edges are finished with sixteen loops of sennit about one inch in length. To 

 two of thfese loops are attached long sennit cords which were rolled up, when the object 

 was pVirehased, dnd thrust into the interior of the cylinder. The cylinder is six inches long 

 and approximately one and three-quarter inches in diameter. After weaving, it seems to have 

 been blacTcened with earth. This record was used as an aid in singing the ■I'avana (Handy). 



-: ■'P'/a'f(? Lx'xxiv, D, shows a geneological record. It is considerably larger 



than B\\\\€ string record just described, and is stuffed with some soft material. 



From the sides spring six long pointed flaps, three above and three below. The knotted 

 cords, . ■rt^b'^served to keep the record, seem to be attached to th'e body without regular order 

 and vary considerably in length. The knpts are all double, qne half of the knot standing for 

 the husband, the other for the wife. The long cords sometimes divide at some distance from 

 the body, and the branches represent other lines of descent. 



The string records, according to Von den Steinen (56), were purely 

 mnemonic and could only be interpreted by the maker, or by one whom he had in- 

 structed. They were used to record geneologies, and also as an aid in chants of a 

 religious nature. Records of both sorts were sometimes attached to a single 

 body. If the record was in the possession of a family, the father made a knot 

 at the birth of each child, and untied the knot if the child died. When a son 

 married, an additional knot was added for his wife. The woven body of a geneo- 

 logical record was called too, signifying the tap root of a tree, and symbolized the 

 history of the gods and the legendary period of the world. The flaps attached 

 to the body are said to symbolize the brothers of the god Atea. 



Deaths may also have been recorded by knot records. Von den Steinen 

 says (56, p. 113) : 



"I was told that in ancient times, a post, hung with many string records, stood on the 

 western cape of the island of Hiva Oa. There the souls of the dead sprang from a high 

 rock into the sea to begin, submerged, their journey to the ancient home, Havaiki. A priest 

 was delegated to register each death which came to his knowledge with a knot." 



[183] 



