352 Memoirs Bcniicc P. Bishop Museum 



The popoi trough is usually made from breadfruit wood. .\u average specimen is 5 feet 

 long and 16 inches wide, with square ends, straight sides, and a rounded lower and slightly ex- 

 cavated upper surface. In new specimens the depth of the hollow rarely exceeds llA inches 

 but in old specimens it may be as much as 3 inches deep in the center, due to wear. In one very 

 old popoi trough, evidently made with stone tools, the interior had been hollowed with fire. 

 The dimensions of individual popoi troughs vary considerably, but those used in Nuku Iliva are 

 on the whole shorter and liroader than those use<l in the southern islands. 



IMPLEMENTS FOR PREPARING COCONUTS 



Coconuts rank next to breadfruit in the Marquesan dietary, but they 

 seem to have been somewhat less imjiortant in pre-Euroi)ean times. They are 

 gathered by hand, without special appliances. The first step in their preparation 

 is to remove the thick husk. This is done by means of a pointed stake, aboitt 

 two inches in diameter, which is fixed rigidly in the ground, point up. To husk 

 the nut it is grasped in both hands, strtick sharply upon the ])oint of the stake, 

 and twisted away from it, removing a shver of the husk, the process being re- 

 peated until only the inner shell remains. The process appears easy, but it calls 

 for a considerable skill in manipulation. Nuts are usually husked where gather- 

 ed, to avoid carrying unnecessary weight.^ • f _ 



Nuts other than the thinnest shelled drinking nuts are opened by a few 

 sharp blows of a stone given along the line of greatest diameter, an expert being 

 able to break a nut into two almost equal halves in this way. The water in the 

 cavity of old nuts is thrown away, and the flesh itself, rarely eaten, is grated to 

 extract the milk. Coconut graters vary considerably in the details of form, but 

 are all made upon a single principle. 



The graters consist of a stool, or seat of some sort, from one end of which an arm 

 projects more or less horizontall\'. At the end of the arm the actual scraper is fastened. In 

 ancient times this consisted of a jjiece of rough coral, or of a strip of shell with the upper 

 edge cut into teeth. Much care is expended upon the seat portion of .some of these scrapers, 

 which appears the more remarkable as the furniture of a Marquesan home rarely includes 

 stools or other raised seats. To use the scraper the workman places himself astride the 

 seat, and rubs the nut rapidly over the end of the arm, with a downward and outward 

 motion. .\ bowl placed below catches the fragments of meat as tliey fall. \\'hen a sufficient 

 quantity of the grateil meat has collected, water is added and the whole mixed to a paste. 



A strainer is then brought into jjlav to separate the milk from the meat. 



This strainer is nothing more than a thick bundle of fiber, about one foot in length and 

 two to three inches in diameter, which is spread out and drawn across the bowl toward the 

 workman, gathering up part of the paste. The ends of the strainer are then twisted in opposite 

 directions, enclosing- the jKiste, and the whole lifted from the bowl and wrung out. the liquid 

 expressed falling back in the bowl. Tlie dry meat remaining in the strainer is thrown away and 

 the process repeated until only a thick creamy white fluid remains in the bowl. Two species of 

 fiber are used for strainers, the commoner is that from the husk of the coconut. The finer strain- 

 ers are made from the fibers of the steiu of a long-leaves plant, which in leaf form and habit of 

 growth resembles large garlic, and from a reference in Jardin (33) is probably Cy perns ina- 

 creilcnia. No attempt is made to separate the fiber from the substance, the stems are simply split 

 into long strips. (This fiber is also commonly employed for kava strainers, but there is a con- 



[9^] 



