POI POUNDERS. 41 



Fig. 26 would do for poi pounding but the}' all lack weight and the face surface is not 

 of sufficient diameter to do well the needed pounding. In Fig. 37 we have a conical 

 niuller made of coral rock (coral sand conglomerate) which is fairly heavy (4 lbs. 12 oz.) , 

 but while it would strike a forceful blow it would not be so easy on the recover, and in 

 spite of the rather roiigh surface would be likely to slip from the hand. In this case 

 the inventive genius of an intelligent people would soon devise the slender stem and 

 knobbed top. I am able to show the intermediate shape when the stem had been 

 diminished for the better clasping of the hand. Fig. 38 shows a very old muller or 



FIG. 36. WOOD AND STUXK I'Ul'NDKKS, CAKUI.INK ISLANDS. 



pounder found in the ruins of an old hciau or temple. It is roughly- wrought and 

 indicates an early age or little skill on the part of the maker. It almost gives the 

 impression of a lump of clay being fashioned on the potter's wheel. It is the onl_v one 

 of this form I have seen. 



Let not my reader suppose tlial I attach mucli importance to this development 

 of the pounders; there is no chronological sequence so far as known, and while it is 

 eas}' to arrange intermediate forms, it must be always remembered that we have nothing 

 bej'ond our imagination to rest upon. We cannot prove that the simple form was not 

 made long after the so-called intermediate for some special purpose. There are no 

 bones of the cave bear or of any other extinct animal with which these stone tools have 

 been found, and except tradition there is no possible help in dating any of the old speci- 

 mens. Tradition seldom meddles with the common implements of vulgar life, and 



certainl}' does not in many of the ones which occupv our attention at present. 



[373] 



