ADDITIONS AND ORNAMENT. 135 
triangle if we consider the design represented by surface, or one and 
two excavated triangles if the design be regarded as incised; in the 
lower line the composites differ in having the subsidiary surface tri- 
angle double, producing in the excavation the series 1-3. In figure 31 
is found a triangle variant in which the ornament is clearly produced 
by the incision, three associated triangles, of which two approximated 
have a common base direction, the third presents its apex slightly 
between the apices of the pair with its base outward, a most effective 
composition. Figure 97 is from one of the two triangle inlays on 
bosses of the club illustrated in Plate VII a. 
The carved basketry skeuomorphs on these clubs show little variety. 
In general, each element of the web is pictured by straight lines, the 
material being a long leaf, and the fact that such a leaf produces the 
effect of parallelism of right lines is clearly apparent on the club illus- 
trated in Plate VII 6, where such a leaf is seen tied about the shaft. 
There are but two groups of the basketry. The former is rectangular 
without septa in figure 35, with septa in figures 36 and 37. The other 
is a picture of diagonal weaving, in its simple type and without septa 
in figure 38, composite and with septa in figures 39 and 40, which 
represent different directions of the diagonal member. 
No inconsiderable time has been spent in the measurement of the 
clubs of this collection; the record is crowded with detail of length and 
girth and grip. In this mensuration there is acquired a mental picture 
of the amount of the incised surface, roughly 150 square feet, minutely 
covered with incisions in which the units rarely amount to a quarter 
of an inch. This is the floorage of a small room. In the preceding 
pages we have considered the distinctive types of this art of the 
savages, some of the types enormously repeated. Now we take up 
in comparison the decoration employing curved lines. In 15 figures 
(43 to 57), and one of these to be rejected as an erratic of Maori 
provenience, we are not dealing with types selected as representative 
of great spaces, but with the individual instances of curvilinear orna- 
ment. ‘Two of these are the ornament of shaft ends in which the circle 
is set by the form of the piece. If one will take a measure and set 
upon the floorage a space just 1 foot in length and a single inch in 
width, he will find before him as nearly as possible the sum of all the 
curvilinear decoration and its relation to the rectilinear ornament. 
There lurks here a most interesting problem in the evolution of 
design. It may not now be possible to solve the problem, yet it is some- 
thing accomplished merely to be able to state it. So far as we are 
justified in drawing a conclusion from this material it is this: Nuclear 
Polynesia has attained to a very satisfactory stage of development 
in the employment of right lines and combinations of right lines for 
decoration; it is scarcely at the beginning of the employment of any 
