EVOLUTION OF THE CLUB TYPES. I2I 
tenon. In 2478 the socket ornament is found only on the central mem- 
ber of the blade socket, the remainder of the unit having been reduced 
to mere ribs. The tenon ornament, together with the same left wing 
of the blade-socket, is clearly treated in twine patterns. In 3361 the 
central member of the blade-socket lacks ornament, the remainder of 
the unit being mere ribs; but the socket spaces are covered with a rec- 
tangular reticulation which extends over tenon-spaces and over the 
blade, quite characteristic of a piece which is altogether a lifeless fol- 
lowing of a set pattern without the slightest comprehension of its sig- 
nificance. 
Yet just as this rude piece supplies the clue to the interpretation of 
the shaft-tenon, so does it afford the explanation of a distinctive ele- 
ment of the construction whose consideration we have postponed to 
the clarification of the socket problem. Following the Admiralty 
Island method, we might postulate the fixing of the socket in both its 
holdings by the use of gum, of which Fiji has abundance, and by twine 
lashings. We have in Nuclear Polynesia abundant proof of the em- 
ployment of these two materials in combination; in this museum cres- 
cent club 3186 d had been so effectively repaired by these two agents as 
to be serviceable for combat and quite as good as new. In the sketch 
of the socket-piece we observe three knobs and find that they appear in 
all three pieces—one in the inner right-hand corner near the shaft, 
the others in the slotted part of the blade-socket, one at the upper 
corner of each of the angles which divide that region into three mem- 
bers. These knobs on one face of each club are exactly matched in 
position by precisely similar knobs on the other face. In the extremely 
illustrative 3361 we find in addition a panel between the upper left 
ribbed angle and the central unit, and in this panel are three pits some- 
what carefully drilled, and these pits correspond with similar pits on 
the other face; furthermore, each of the three large knobs on each face 
catries a similar pit. Additional to these drill-marks, which so corre- 
spond on one face and the other that they might be the two extremities 
of a perforation, we find on the small remnant of the central solid unit of 
the socket-piece and upon the central member of the blade-socket drilled 
pits ordered in straight lines and quite as distinct as are the others, the 
sole reason for setting them apart in independent consideration being 
that they do not exactly correspond in position on the two faces. The 
only meaning which it is possible to attach to these prominent knobs is 
that they represent from the stone prototype pegs which served to 
anchor the several parts of the combination, the upper single knob 
spiking the shaft socket to the tenon, the lower knobs similarly spiking 
the blade within its socket. Particular significance attaches to the 
upper panel in 3361 with the drill-mark suggestion of perforations; they 
quite confirm the spiking suggestion. I can not now recall, either from 
experience in the life of the South Sea or from collections of ethnica, 
