Linton — Tlic Marquesas Islands 363 



One specimen was described as having a very short loop of cord passed through a large ring 

 of carved tone, which served for suspension. In very large feast bowls the handle was made 

 of a heavy plaited cord of coconut fiber whose ends were attached at two points, four to five 

 inches apart. Such handles were usually encased in a series of bone tubes, made from the 

 small bones of the lower arm or leg. These were similar to the tubes used on the cord handles 

 of coconut shell bowls. The native name of these tubes was not ascertained but according to 

 Handy (32) Dordillon gives a term, vipoo, which he defines as "a piece of bone to ornament 

 dishes or drums." It seems probable that this refers to the small carved figures rather than 

 the simple corrugated rings, for the rings were not used on drums so far as could be learned. 

 It is rather interesting that the large feast bowls had handles on the side only, indicating 

 that the handles were used to suspend or drag them, not to carry them. 



Wooden handles, cut in one piece with the bowl, are very rare in the cir- 

 cular forms — onl}^ three examples were found. 



In one of these specimens there are four small, equally spaced projections a short distance 

 below the rim. These projections are carved into peculiar little double figures, and as the rest 

 of the bowl is uncarved the projections are probably intended for ornament, although they could 

 have been used for handles. 



Another bowl seen in a native household, had a flat handle about six inches 

 long projecting from one side of the rim. The shape of this handle was 

 identical with that of an ordinary American frying pan, and it seems probable 

 that this development was due to European influence. Still another bowl (PL 

 Lix, E) now in the Bishop Museum, is tgg shaped rather than round. 



It is a rather deep well-made bowl, 9 inches wide and 10^^ inches long to the base of 

 the handle. The handle springs from one side of the rim and is Z]4> inches long and an inch 

 wide in the center. The outer end is expanded both laterally and vertically, and the tip has 

 a deep vertical notch. The whole handle has very much the shape of the distal end of a 

 human humerus, and may very well have been modeled after one. 



It seems probable that covers were employed with some types of round bowl 

 other than the hue po'o just described, but if so their use has been discontinued 

 for many years and no examples have been preserved or described. 



Some round bowls, especially those of recent manufacture, are provided 

 with low circular bases. A similar feature is found in certain tureen-like oval 

 forms (p. 365). Most of the natives believe that this feature is a recent in- 

 novation, copied from the Chinese rice bowls introduced by European traders. 

 On the other hand, a very beautifully carved bowl in the possession of Mme. 

 Mallius, of Ta'aoa, Hiva Oa, shows this feature, although it is obviously an old 

 specimen. It seems probable, therefore, that if the circular base is non-Marquesan 

 it was introduced during the early period of contact. 



The transition from round bowls to the oval and rectangular forms ap- 

 pears to be a gradual one, every step being represented by a few examples. No. 

 I shown on Plate Lx, A, is especially interesting in this connection as it has all 

 the featfires of an ordinary round bowl except that the rim is slightly flattened on 



L103] 



