346 Memoirs Bcniicc P. Bislio/' Muscii))i 



Figure B, on the same plate is also highly conventionalized, but it ap- 

 pears less grotesque. It is of nearly the same height, but the figure is more 

 slender and the proportions more nearly ai)proach those of the human body. 



The head is large and is set directly upon the tiody. but the mouth is smaller than in A 

 and the nose is rather small and naturalistic, with a high briflge. Tlie eyes are shown as long 

 ovals having a raised rim and central ridge, and are connected with the nose by narrow bands. 

 Above each eye there is a raised band whose imier end connects with the base of the nose 

 while the outer end, behind the eye, thickens to form a small ear flange. The arms are 

 carried in the same position as those of .4 and are unmodeled, but the projxjrtions of upper 

 arm and fore-arm are almost natural. The fingers are not indicated. The body is jjropor- 

 tionally much longer than that of A, its sides are rounded, and the whole treatment is sur- 

 prisingly naturalistic. The back of the body is nearly straight. .Along the middle of the top 

 of the shoulder there is a well marked ridge running from the head to the shoulder tip. From 

 the tip of the shoulder a second ridge slopes downward and forward across the chest, its inner 

 end curving upward abruptly toward the throat. This probably represents the line of the collar 

 bone. A short distance below this line there are two small smoothly rounded protuberances 

 which represent the breasts, while the navel is shown by a small circular pit placed midway 

 between the hands. The sex, male, is indicated by a small conical knob. The short legs are 

 flexed, but the thigh is rounded and the buttocks project to only a natural degree. The lower 

 leg is thick and heavy, and the foot is large. The ankle bone projection is well marked. The 

 back of the head was pierced for suspension by boring two holes, one downward from the top, 

 and the other upward and forward from the bottom. 



The convention of the double figure ( PI. Lv, C ) corresponds approximate- 

 ly to that of the single fisfures iust described. 



'te^ 



The two figures are placed back to back and joined at the buttocks and at the shoulders 

 or backs of the heads. The form of the head differs somewhat from that of the other two fig- 

 ures described, and one figure has a hand to the mouth, a pose not uncommon in Marquesan art. 



A curious object shown on Plate i.v, D, which may have been a votive of- 

 fering or have represented a deity, was found in an old iiic'ac. 



This article is a roughly conical piece of cellular lava which appears water worn and 

 shows no signs of working except at the upper end where it has been carved into two heads, 

 set back to back and se])arated by a groove. The stone is so coarse that it is impossible to 

 determine the convention of the carving. The lower end of the object is smoothly rounded, and 

 but for its size and weight the whole might be taken for a very crude f'opoi pounder. Natives 

 could throw no light on its use. 



Stone figures appear to be entirely lacking in Samoa and Tonga, but small 

 figures are found throughout most of marginal Polynesia. They are numerous 

 only in the Marquesas. Fish figures are found in Hawaii, the Marquesas and 

 Society Islands, although there is little information as to their use and significance 

 in Hawaii. Small human figures occur in Hawaii and the Society Islands, but are 

 rare in these localities. The little kumara gods of New Zealand are well known 

 and appear to be fairly common. The figures from the different localities seem 

 to conform in every case to the local convention and have little in common. 



[86] 



