So 



Ka Hana Kapa. 



I have a photograph of a beater in the British Museum, apparently from the Society 

 Islands, with the ridges not only rounded in transverse section, but neath" rounded 

 to a point at the distal end. I am inclined to consider the pcpclii the older form ; it 

 would be easier to make, and is found commonh- on the beaters of other groups uhere 

 the art had not reached the level of the Hawaiian. In the oldest the ridges are flat 

 and rather wide, and there are seldom more than five on a side of the beater. Later 

 the ridges became rounded and closer together until they could no longer be finished 

 with a round lop and were of the shape of, and, in fact were lioopai. In Fig. 31 may 

 be seen several degrees of the pcpclii form, while one of the finest specimens of the 



v- -".i'.n 1 



i/fUiy/./'iii'.'yi'.-- 





iitfif'J'//'.'" 



241 



199 



203 255 189 



Fig. 34. Forms of ie kuku. 



8673 



/ioopai\s shown in Fig. 34, No. 6. It will be remembered that all the markings on 

 the preliminary beaters or liolioa were of the Iioopai form, but generally coarser and 

 deeper cut than in the specimen of beater just referred to. We will now turn to the 

 less common and more elaborate designs. 



All the.se had some .significance, as shown in their names: thus kocaii seems to 

 come from Xw, an earth-worm, and an signifjnng motion. Fig. 34, No. 5. This name 

 applies when the undulating ridges are parallel, but when they are not so, but arranged 

 as in Fig. 35, Nos. 9 and 10, the name \s piiiIi — 3. twining. The two are often com- 

 bined in various ways as seen in Fig. 34, No. 2, and there is one specimen in this 

 Museum (No. 205) where the two patterus alternate in sections a little more than an 

 inch wide down the face of the beater. We will presently come to other modifications 

 appU'ing to all the patterns already mentioned, but we will first note the few other 



