i6o 



The Ancient Haivaiian House. 



the specimens show, Avhat can be seen in those figured, that the rough block was only 

 approximately circular, and it seems certain that the final rotundity was given by the 

 accurate eye of the artisan. Whether the result of warping or not, the accurate meas- 

 urements of the genuine old umeke always detect a small deviation from the circle 

 which can sometimes be seen by the ej-e. Thej- however used for preliminary meas- 

 ures a flat strip of bambu which also served for a "straightedge" or rule: I have 

 never seen one of these bambu rules graduated or marked in an}- definite way. 



In excavating the inside it is curious to note how they adopted the method of 

 the modern turner as shown in the first bowl in the figure, where an excavation was 



FIG. 141. OI'AKA OR POLYHEDRAL UMEKE: 523-488. 



carried to a certain depth then an inner concentric circle was dug out in proportion to 

 the outside curvature, and finally the angular benches were dug or grated down to the 

 final surface. In all cases where a handle was desired a part of the block was left for 

 that purpose, a thing the turner cannot do: he must carve his handle in a separate 

 piece and glue or pin it on. In some of the old ipu kiilia or spittoons found in burial 

 caves and supposed to be of considerable antiquity this handle is of slight projection, 

 in others has been bored for a string at some time subsequent to its original making, 

 for the bore is rough and of inferior workmanship: these handles can be compared in 

 the plate where many ipu kuha are photographed. (Plate XXXII.) 



In some of these half-made bowls, I think I can detect the mark of a small, 

 sharp adze, and again the marks of a stone chisel, but the tool which seems to a mod- 

 ern amateur carpenter most efficient, the shark tooth cutter, does not appear to have 

 been used on these blocks. It was, however, the favorite tool of the carver of figures. 



[344] 



