ADDITIONS AND ORNAMENT. I4I 
sea-birds dive into the water for fish; there is only one which chases 
it along the surface, as established by the interrupted line. This is 
the triple play of bonito, flying fish, and albatross; the bonito under 
the surface drives the silvery and toothsome fish, which takes to air in 
its gliding flight, and there stands an equal chance of being snapped up 
by the master of that element. ‘The sight is frequent; it would natur- 
ally suggest itself to the observant artist; we may be warranted in 
reading into it a valid club suggestion, for the food motive was never 
very deeply buried beneath the surface of combat in these islands. 
The bird series begins at figure 75 and includes figures 90 and roa, 
the first six being quite graphic, at the other end highly conventional- 
ized; but the two very effective groups of flight serve conclusively to 
establish the convention (figs. 81 and 82). Specific characters are 
very scantily indicated in this collection. © 
In figures 75 and 76 we feel warranted in the belief that the same 
bird is portrayed, despite some slight differences in the execution, a 
length of bill and of tail accompanied by straightness of legs being 
similar in the two carvings. In figures 77 and 78 occurs a common 
character in the triangular form of the legs; I interpret this as a con- 
vention indicative of the web-foot of the sea-birds and shall undertake 
to support the principle of perspective when we discuss the far more 
important drawing in figure 142. Sufficient attention has already been 
given to the birds in the figures 79 to 82. The remaining figures are 
all most highly conventionalized, a plumage distinction being at least 
indicated in 86 and a peculiarity of head in 90 and 104. Six figures 
remain in which the bird is represented by a generally similar design 
of five angles, which differs from the five-pointed star of our decora- 
tion by the consistent absence of the reentrant angle at the base of 
that design, a feature which seems to represent the tail of the bird 
beyond any doubt. In figures 85, 87, 88, 89, 90, and 104, the rear 
line of this tail is remarkably straight; in figure 83 it is a considerable 
are of a circle; in 84 it has been mutilated, but there persists somewhat 
more than a suggestion of such curvature. 
Figures 91 and 92 might readily pass for extremely formal and pre- 
cise drawings of the five-pointed bird design. Yet on the advice of 
Samoan commentators I set these apart as pictures of the octopus, and 
we need such explanation in support of the series of derivative forms in 
inlay which are wide of the bird suggestion. The octopus is a sac, a 
webbed disk, and tentacles. Now, if one holds an octopus by the sac 
and lowers it in the air to a plane surface, as it has been shown to me on 
dry beach-sand, the tentacles retract beneath the webbed disk and the 
horizontal profile tends to approximate a more or less regular eight- 
pointed star. If now the sac is lowered and the support of the hand 
removed, it tends to flatten out, because the consistency of the flesh is 
not sufficient to support all of its own weight when out of its element; 
