DIMENSIONS AND STRUCTURAL DETAILS. 89 
of wood from which he is to carve a club; he grasps it in his fingers 
with the end approximated to his palm; he stretches it across his body. 
and marks the point upon the stick that corresponds with the particu- 
lar corporal unit which custom has decided upon for the length appro- 
priate to the particular type of club which he is to carve. In taking 
these control measurements under conditions which have come under 
observation in the islands, it became at once apparent that there is 
physical reason for the variety in the magnitude of this factor which 
is constantly applied; it is noted that the billet employed in these 
tests was a billiard-cue held by the butt with a circumference of 
5 inches. When we attempt to adjust the elbow-measurement to the 
list of club-measurements, we note that its maximum of 19 inches is too 
great and its effective measurement of 
14 inches is too small. The clubs to 
TABLE 39. 
which this unit might apply are all of Inches. __|Pieces. 
the ula type—a large ball with a slim 
handle, carved from the root-knob of a | Etbow unit...| (19-14) 18 
sapling. After trimming up the root- : age ag 
Fer “ Sternum......| (36-30) 38-30 29 
lets it is natural for the clubwright to | Contra-axilla .| (43-39) 44-39 | 71 
hold the ball in his hand and to take | Contra-elbow.| (53-46) 53-45 | 31 
the stem measurement up his arm as 
far asthe elbow. But the clutch upon the root-ball is quite other than 
that which he will employ when holding a somewhat heavy billet 
without the support of the forearm which will be effective in the case 
of the lighter ula sapling. The majority of the ula are found in the 
sixteenth inch of length; test of the specimens has shown that this 
length is exactly offered in the fashion in which the clubs are held for 
measurement toward the elbow. 
The reduced handing measurements are not in the least prohibitive 
of the maximum measurements of each unit, for clubwrights have 
been observed measuring their raw lumber when held in the hand of 
an assistant or apprentice. We are justified, therefore, in taking the 
two values of each unit as limital. Thus we are able (table 39) to 
classify the clubs by linear units of this corporal sort. 
Thus we see that all of these clubs are governed by a generally 
established system. ‘The small and very elegantly executed mace 
(3792 b) is anomalous; it is the shortest of all these pieces (14.5 inches), 
yet it has a haft-girth corresponding to a length and weight much 
greater. Because of its small size and evident effectiveness, one is 
tempted to speak of it as a pocket weapon until it is recalled that 
costume in the Pacific has not yet arrived at marsupial convenience. 
The next group of measurements which we shall examine deals with 
circular measure in its earliest phase, for the relation of the radius 
and the circumference are unthought of as yet and the constant 7 lies 
ages distant. The savage of the Pacific measures such circular dimen- 

