Li2 CLUB TYPES OF NUCLEAR POLYNESIA. 
a structural absurdity, it seems to be clearly of the same type as that 
which we have examined in the New Guinea group (Plate V, 13 to 16). 
Regarding this as a movable addition, we shall try to discover what 
remains in the way of shaft after such removal. We find a smoothly 
worked shaft of fairly even diameter, except just at the bend, all as 
set forth in table 48, where we have pointed out the gradual increment 
of an inch at bend and head. ‘This shaft uniformly ends in a plate, 
commonly guttered very slightly, from which rises a cone generally 
of sufficient height to afford a somewhat sharp point. We find in the 
Pacific no such cone-headed pieces with a bend, but we do find pieces 
in every other respect the same, barring the curve. Parkinson (Tafel 
8, figures 7 bau and 10 mukmuk) pictures from the Gazelle Peninsula, 
but as probably ferried across the strait from New Ireland, straight 
clubs which end in a distal plate and sharp cone, these being straight 
from haft to head. He presents from the Sulka and O Mengen (p. 229, 
figs. 4 to 7), but probably derived from the mountaineer Baining, clubs 
of straight shaft, which end in distal plate and sharp cone, and, this 
being a particularly important detail, have carved for some distance 
inside the distal plate an array of knobs quite closely resembling the 
stone heads from New Guinea. We shall then have to conceive it possi- 
ble that the plate-and-cone shaft has been fitted with the stone head, 
that in the course of migration the present people of Fiji have acquired 
the hockey-stick shaft of Arossi, as seems confirmed by Niué, and 
that under this intermediate influence the straight pandanus club of 
the Nakanai coast of New Britain, which itself is now an all-wooden 
metamorph, has become the curved pandanus fotokia of Fiji. In this 
case we postulate the secondary evolution when the Fijians have 
recognized the similarity to the pandanus-fruit cluster and have carried 
it still farther in the botanical detail of nutlets radiant from a common 
core (Plate VI, fig. d). 
We interrupt the study of the curved clubs in order to complete the 
tale of the spiked stone head which we have been following through 
various mounts. ‘Two of the maces (Plate III, figs. b and c) are clear 
developments of the same head-theme. ‘The larger of these pieces is 
of poor workmanship, the other most artfully worked out. They 
exhibit a straight shaft, a spinous head with attachments distal and 
proximal. In the larger piece these are represented by merely whittled 
cone attachments, but in the smaller we find the end capped by a 
neatly worked plate slightly domed and at the proximal end of the head 
a small circlet of lighter spines in a contrary spiral. The distal plate 
or cone we can readily interpret as a finial of the wooden shaft 
intended to prevent the stone head from slipping off in that direction. 
It appears probable that the proximal attachment represents some 
arrangement of sennit lashing of or gum; we have seen each material 
employed to this end, devised to hold the head secure upon the shaft 
