William Churchill.

Club types of nuclear Polynesia online

. (page 12 of 21)
Online LibraryWilliam ChurchillClub types of nuclear Polynesia → online text (page 12 of 21)
Font size
QR-code for this ebook

that the pentagonal and square lugs may evolve from slips in the carv-
ing or from trimming up lugs which have been shattered by the acci-
dents of use, gentleness not being particularly characteristic of club
play. In the carinated club 2285, we find a most unusual position
for a lug in this province ; in addition to the lug in the usual position a
second is carved upon the shaft near the head.

Table 45.

Semicircular lug.

Pentagonal lug.



Samoa, 3172 a, 3172 6, 3178 a, 3789

Samoa, 3099 a, 2283

Tonga, 3355. 3359. 2261, 2260


Plane :

Samoa, 2278, 2281, 3178,


Samoa, 2272, 3788, 2280, 2273, 2276,

Fij', 3147 a, 3100 a


Square lug.

Tonga, 3173. 3174- 3358


Triangular lug.

Samoa, 2270

Vertical :

Tonga, 3356

Samoa, 3788 a, 3099, i5743. 2284

Tonga, 2259, 3357. 2264, 2271, 2256


Samoa, 2274, 2279, 2287, 2285

Tonga, 2257

The perforation of the haft-end shows 3 types which extend over
Nuclear Polynesia. The simplest is merely the drilling of a hole at
right angles to the face of the lug and parallel with the surface of the
club-end; this is found in table 44, under the designation "Transverse."
Its recorded frequency in Fiji is immaterial, for there are but 2 lugs
in that group and both show this simple perforation. The second
perforation in frequency is the V-type, in which the drill is not set
parallel with the end of the club, but two perforations are made at
equal angles on each side of the central point and continued until they
meet below the surface, and in many cases the perforation thus made
has been rubbed down with a thin strip of the skin of the ray until the
piercing approximates the semicircular. In Samoa the V-perforation
occurs in all 7 instances with the lug; in Tonga 4 times with the lug
and once without; in Fiji all 3 instances are without the lug. The
diagonal perforation is essentially apart from the lug; it consists of a
hole bored from the haft-end at such an angle as to come out upon
the haft near the end. This perforation is found in vSamoa in but 2 out
of 29 perforated clubs, in Tonga in i out of 20, but in Fiji in 5 out of 13.


The increase in the proportion in Fiji is not to be taken as critical of
that culture; it results from the paucity of the lug in that archipelago.
Fiji supplies a perforation that is found nowhere else, the inverted V;
in this, two holes are drilled from the side of the haft toward its end
and at such an angle that they meet and issue from the end as one ;
that the drilling is in the direction stated is estabUshed by the ula
3786, in which one hole has been completed and a second hole has been
started on the side of the haft to meet it. Fiji also affords two
instances of the perforation of the shaft diametrically, undoubtedly
a foreign contamination.

The purpose of the perforation is apparent from observation in the
field and from the several museum specimens which still retain an
original becket of sennit. This becket was solely for the purpose of
suspension of the weapon; it partook in no sort of the nature of the
sword-knot designed to be caught over the wrist as a protection
against disarmament.

The single specimen from Nine, iiluhelu, 18094, is of a peculiarly
interesting haft type. Outside of the grip the haft-end is a cone of
9 inches length and diameter of a quarter inch at the tip, which evenly
increases to a circumference of 4 inches at the base, and there is finished
off with a raised ring of 5.5 inches circumference, and immediately
follows the grip with a sharp reduction to a circumference of 3.5 inches.
It is manifest that this haft-end serves no end of utility, except that
the raised ring gives security against a possible attempt at disarma-
ment by pulhng; as we have noted in the case of the flange and knobs,
the cone is purely decorative. We find nothing at all like this in
Nuclear Polynesia, but the motive is frequent in the club culture of
Melanesia and may be recognized in Parkinson's "Dreissig Jahre,"
from the Gazelle Peninsula of New Britain (p. 112), from the Sulka
and O Mengen (p. 229) of the same island — an interesting fact when
we note that the shape of the blade is found in the Solomon Islands
at San Cristoval (Revue d' Ethnographic, 1885; T. Verguet, Arossi;
cf. figs. 13, 29, 30).

The shafts of clubs offer less variety in treatment. In general they
are cylindrical, with certain varieties of girth at various parts of length,
as will appear on consideration of the record of dimensions.

A cross-section generally oval appears in a few pieces, all of Fijian
provenience, as listed: billet, 2488, 3184; axe-bit, 2478, 3361, 3362;
serrated, 2496, 3790 h.

Hexagonal section is found in 2 mushroom clubs from Samoa, 3789
and A 15743.

The Tongan paddle clubs have generally a circular section at the grip,
becoming oval as the blade is approached; the same is seen in the
dancing paddle from Fiji, 2501.

Two maces from Samoa, 3792 and 3792 c, are circular at the grip
and become square near the head.



A talavalu from Samoa (3788) and a crescent club from Tonga
(2500) have circular grip and become rhomboid near the head.

A rootstock from Tonga (3175) and an ula (2465) are circular and
hexagonal at these two points.

Three Fijian rootstocks are circular at the grip and develop flat
planes near the head in alternation with the flanges of the head ; these
are 2482, 3100, and 3782 a.

In certain types of club one or more angles persist either from a
recollection of the original article conventionalized in the club form,
as the lapalapa from the angled coconut leafstalk, or ribs or angles,
which are specific in the head form, extend more or less down the shaft,
as in the carinate, the crescent, and the serrated types. In general,
these angles are smoothed out in the final finish of the piece when it
is rubbed down w^ith shagreen or abrasive stone, or if they are allov/ed
to remain they are smoothed off in the grip for convenience in hand-
ling. These are noted in the following memoranda :


Angles persistent throughout, 2270, 3172 a, 3172 h, 2277.

Angles stop at grip, 2280, 2281, 2278.

Angles stop near head, 2276.
Crescent :

Angles persistent throughout, 2263, 3186 d.

Angles stop at grip, 2284, 2499.
Serrated :

Angles persistent throughout. 3791 c.

Angles stop near head, 3790 h, 3790 a 2497.

The billets are the only clubs in which it is possible to study the
difference between haft and head dimensions with anything like the
detail which has been possible in the increment from the grip to the
haft-end, whether as flange or knob. The moment of such increment
in circumference, by inches and fractions, appears in table 46.

Table 46.


Piece No.

(inches) .

Piece No.


1 25



3177. 2491, 2488, 2489,3100 b,

3100 a, 3780 c, 3147
3780 a

2492. 2490, 3780 d, 2493, 3780
3186, 3144
3147 a






3184, 3780 <;





There seems to be no connection between these forms and the source
of the pieces, and the increment does not at all correlate with the
length of the weapon.

In all other types the head is little related with the shaft, and in the
head variety there would be no profit in comparison of dimensions ; yet



a comparison is possible among the shaft dimensions for different

In the rootstock type the shaft and head are regarded as continuous
in source in the simpler form. The dimensions of shaft at grip and
at the nearest point to the head give the increments shown in table 47.

Table 47.

(inches) .

Piece No.


(inches) .

Piece No.

1 -25

3782 a

3783, 3782 b, 3100. 2482, 3782

3303 a. 3782 c. 2479

3175. 2485, IQ74. 2484



The missile ula presents two types of shaft, which will eventually
be found charged with significance. All of the ball-headed type and
almost all of the flange-headed type have shafts which are of uniform
diameter from haft to head. Two of the latter form, 2465 and 3786,
show a swelling of the shaft toward the head amounting to 1.75 and
2.25 inches of circumference increment respectively. I incUne to
regard this as influenced by the form of the next type. All of the type
with the patterned head show increments: 2466, 0.5 inch; 3188 a,
1.5 inches, swelling into the pattern of the head; 2461, 1.87 inches;
3188, 2 inches; 2461 a, 1.25 inches, and a distinct plate at head;
2462, 2.75 inches, and similar plate.

Table 48.




Piece No.

at haft

at bend

at knob

(inches) .


(inches) .

3182 c
















3182 b





4 5











The pandanus clubs show clearly a shaft design in various stages of
execution. In 5 of the clubs it is manifest that at the three distinct
points of the shaft— the haft, the bend, and the knob— there is intended
to be an increment of an inch. In the next two the dimensions at
bend and at knob remain unaltered ; in the last there is a sHght decrease
at the knob. These will be apparent on comparison of dimensions at
the critical points, shown in table 48.

The three axe-bit clubs give these dimensions of shaft: 4 inches,
5.25 inches at head, flanging to 6.5 inches; 2478, 4.25 inches, flanging



to 6.75 inches; 3361, 4.75 and 6 inches, no flange. These pieces lack
the scrupulosity of the work of clubwrights working in other types,
yet none the less on that account is every detail significant. There can
be no doubt that the flange is integral to the shaft and a necessary
part of the design. Piece 3361 is a very rude specimen; the flange
escapes the tape, but its motive immediately appears when we examine
the next unit of the head, for there is a notable decrease in perimeter,
implying a flange.

The lipped clubs off'er little of interest in their shaft detail. In two
of the types the shaft is of uniform diameter as far as the beginning
of the roughened panel, and there engages with the detail of the head ;
in the highly ornate type with worked panels the diameter of the shaft
increases slowly from haft to the same critical point.

Table 49.



Piece No.

Ends similar:







3147 a, 2267. 3100 b, 3100 a, 3780, 3780 a
3147, 3780 e


3144, 3185, 2491, 2488, 2490, 2489, 3186, 3143. 3780 c

2265, 3177. 2492, 3780 d, 2493


Ends dissimilar:




The maces have shafts of practically uniform diameter; probably no
significance attaches to the fact that in 3792 a the shaft circumference
immediately next the head is half an inch smaller than at the haft.

In the talavalu the shaft is generally of uniform diameter, such
enlargement toward the head as may exist being entirely due to the
rather distinct shoulder from which arises the serrated blade; in 2272
this expansion gives the dimensions of 4.5 inches at haft and 11 inches
at shoulder.

The lapalapa follow the general shape of the leaf-stalk from which
they have been conventionalized — a smooth increase from haft to head.

Similarly the paddle clubs, offering no distinct demarcation between
shaft and head, do not call for consideration here.

In the carinated clubs the shaft is of equal diameter throughout
until the point is reached where it expands toward the head unit. The
same note applies to the serrated, the crescent, and the mushroom clubs.

In the 2 horned clubs measured the shafts are generally of the same
diameter; one increases by a slight amount next the blade, the other
similarly decreases.

The head-forms are so various, congruent in each type of weapon
but not susceptible of coordination, that it will be just as well in this
place to indicate the separation in source, which will be subjected to



argument in the next chapter. The clubs which without any per-
adventure are derivative from a strictly wooden source are the fol-
lowing: the crescent, the billet, the rootstock, the lapalapa, the paddle,
the serrated, the mushroom, and the carinated; of these the billet
alone affords any possibiHty of coordinating the two extremities of
the same piece.

These fall into two groups of somewhat unequal numerical value.
In one, with 8 pieces, the two ends are similar; in the other, with
14 pieces, they are dissimilar. These are listed in table 49.

With such diversity in a weapon of such simple form it is clear that
the variety lacks significance; the only conclusion which we are justi-

Table 50.


Piece No.


Piece No.

II. 9

2257 a


2271, 2269




3146, 2261


3174, 1975. 3174 a. 2259


3359. 3145. 2260


2264, 3360




3357. 3355




2268, 3356





fied in drawing is that a marked preference has been shown for the
domed head. It is interesting to compare with this end-treatment the
interesting end of the rootstocks, which consists of a well-formed
cylinder ending in a cone or cut square across and without trace of
the dome motive. This will arise for consideration in the following

The crescent, mushroom, and horned clubs have received sufficient
consideration at their first presentation. We can discover in their
varied forms no suggestion of origin other than derivation from the
possibilities of timber.

The lapalapa at the head-end carry, as already stated, the suggestion
of the cupping which is found in the natural leaf-stalk from which
they derive. This character is so fundamental that it is observed in
II of the 17 pieces — 2270, 2280, 2281, 3178, 2279, 2278,3172 a,
2277, 2276, 2274, and the interesting 2273, which exhibits the rare
and imitative unit of a strongly carved band over the middle of the
cupping. Five pieces lack the cupping and finish with a more or less
straight edge — 3099, 3172, 3173, 3172 b, and 2266. One only —
3178 a — has so far departed from source as to present a domed end.

The heads of the paddle clubs vary considerably in shape, though
not at all in design. This form depends upon two factors, the maxi-
mum width and the distance of that width from the tip of the head.


In an empirical fashion I have gathered these variants into groups for
comparison by extracting a definite index-figure as the result of divid-
ing the width by the distance. We then find one large group in which
the width exceeds the distance, a smaller group of 3 pieces in which
they are equal, and a final group of 3 pieces in which the distance-
exceeds the width. Assembled by these empiric indices the clubs of
this group fall into the array shown in table 50.

These complete such notes as are needed in the record of the clubs
whose origin is distinctly from the wood employed. We next proceed
to the consideration of the club types in which the form and struc-
tural detail of the head and parts thereto adjacent are insusceptible of
comprehension as timber products.



-1 1 I I ] I I ■

Metamorphs of Club Heads; Series A.





e f

Metamorphs of Club Heads : Series B.



It is incumbent upon me to present this theme with great nicety of
detail in order that the reader may be led in the direction of the con-
clusion which has been forced upon me in the intricate task of ordering
the clubs by types and of studying the meaning of every unit of struc-
ture as the clubs passed through my hands.

In the study of the actual weapons of the types to which I have
assigned the designations of ida, mace, talavalu, lipped, pandanus, and
axe-bit, I have convinced myself that we have to do with wooden
metamorphs of similar clubs in remote prehistory in which the effective
head was stone or shell hafted in wood. This conviction I hope to be
able to communicate with the aid of the illustrations and of the detailed

At the time of discovery by European navigators. Nuclear Poly-
nesia was found in possession of stone utensils, but we have no record,
nor have the islanders themselves any tradition, of the employment
of stone-headed weapons. The adze was in constant use, a mass of
hard volcanic rock, polished, worked to a cutting-edge, and mounted
with its edge transverse to the wooden handle, to which the stone was
applied with great ingenuity, the summit of this art being preserved
in the museum in the several ceremonial adzes deriving from Mangaia
and Raro tonga. Minor edged tools were subsidiary to the adze, stone
chisels, drills, scrapers of various uses, and particularly the some-
what highly speciaUzed scraper employed for the shredding of the
dense meat of the coconut in order to extract therefrom in combina-
tion with the water of the nut, that emulsion which enters so largely
into the island dietary and has become known as the milk of the coco-
nut, a thing quite unaccountable in the popular saying, because the
milk never is in the coconut, but is a product of the ingenuity of man.
The extent to which stone cutting- tools were in employment is indicated
by the fact that all our dictionaries of the region afford us the word
foanga, or some dialectic form thereof, with the definition of grind-
stone, a misnomer, since the stone was not rotated, but served as a
rubbing-stone or whetstone. Thus, while we are fully assured of the
employment of the hafted stone blade as a domestic and industrial
implement, we have no knowledge as to the use of hafted stone in

In this lack. Nuclear Polynesia is set apart from the later culture
of its dominant race, for in the regions of settlement by Tongafiti folk
we find abundance of stone armament. It is equally set apart from
all the tangled Melanesian cultures to the westward through which
we postulate the leisurely migration of the Proto-Samoans upon their



course to the settlement of Nuclear Polynesia. The Tongafiti Poly-
nesians, all the Melanesians of whatever culture horizon, have weapons
of stone; the Proto-Samoan population of intervening Nuclear Poly-
nesia alone show this lack. In the prosecution of this particular
inquiry I hope to establish my conviction that in the particular types
mentioned we have wooden weapons which show the manufacture in
the more readily workable material of forms which at some remote
time, and perhaps place, of origin Vv-ere more painfully worked in stone
or other hard material possessing armament value. Wood is more
readily worked than stone, yet the wood employed for weapon pur-
poses, largely the very heavy and dense ironwood {Casuarina equiseti-
folia), is practically indestructible, save by the accident of fire, as will
be apparent when it is recalled how few are the occasions in the detailed
description of the pieces when it has been necessary to comment upon
any club as worn or broken. Despite the intimacy of daily acquaint-
ance with these clubs, continued through many pleasant days, I
myself can now recall in such memory in the list of breakages a pan-
danus, a mushroom, a paddle, a talavalu, and a crescent, five in all,
and of the five but one in which the breakage could seriously impair
the value of the weapon.

We postulate, therefore, that wood is appreciably more easy for the
clubwright to work with his stone knife; that it is quite as durable
within the circle of utility for the specific purpose of ofTense.

In the inquiry upon which vve now are to enter we shall look for the
evidence that these are wood metamorphs upon stone to be presented
to our view in three principal points : the shaft of the club at the head
is to offer, either in its distal or in its proximal aspect, the proof that
shaft and head are distinct entities in theme ; the head is to show trace-
able resemblance to head-types which we can discover in stone in regions
or upon culture planes with which it is demonstrable that Nuclear
Polynesians have come into contact; the third important detail, one
not always present but extremely significant when it does exist, is the
carved band which engages with the head and some portion of the shaft.
A fourth significant point, this restricted to the lipped clubs, is a dis-
tinct edge. With these two postulates and with the attention directed
upon these four critical points, we now take up each type of club in
which we find metamorphism.

The easiest of approach, because its form is the simplest, is the ula
or missile club. Even in this simple type we have found three species
set apart by the form in which the head is treated. The least com-
plex species is that in which the head is merely a ball smoothed as much
as the natural root-ball will permit. Of this type we have 5 pieces,
2468, 2469, 2467, 3785 a, 3785. In each case we find the treatment of
head for which we have adopted the formula "saucered at shaft." In
close approximation to the point where the wood of the head and the


wood of the shaft unite, the head loses its spherical curve and is carved

in a shallow depression which suggests nothing so much as a saucer,
and from this saucer the shaft arises. This is visible in Plate V, figure
12 (2468), the same piece as is shown in Plate I, figure c. In this illus-
tration is observed a very shght increase in the diameter of the shaft at
the point of union with the head. The fact that this increase is slight
and is not found in companion pieces leads us to lay aside this detail as
not affecting the general problem ; possibly it is influenced by the next
higher species. Although these pieces are all carved in solid wood,
each looks as if the handle had been driven into a hole drilled part way
into the head for its reception. We find no stone-headed club in the
neolithic culture of the Pacific which exactly carries out this suggestion,
no piece in which a stone rubbed down to a more or less accurate
sphere is pierced by the drill partially through its diam.eter. If we
were to find such a stone thus partially perforated v/e could compre-
hend how a shaft might be driven into the hole, chocked in place by
subsidiary splints, and finally secured by the application of gum be-
tween the shaft and the edges of the stone, which would take a conical
form through the operation of grinding the perforation. Even though
we lack this stone head, we feel justified in establishing it by extrapola-
tion upon the series of data which we possess. Immediately following
this hypothetical sphere of stone half-perforated comes the spherical
stone complete^ perforated. The museum affords us an abundance
of specimens of this advanced stage. In Plate V, figure 11, we have
an excellent specimen of such a stone removed from its shaft and exhib-
iting the stone made spherical by rubbing, the perforation along the
diameter passing quite through the stone mass, the conical rim of the
perforation which pairs with the similar depression on the other face
to produce an hourglass section. In Plate V, figure 17, we show the
other face of the same stone club-head, which still contains the gum
which has filled the conical depression for the purpose of anchoring the
head to the shaft, this gum being molded into form such as will com-
plete the curve of the sphere up to the handle, as will be seen in figure
18, a complete weapon of this type in which the head is so finnly at-
tached to the shaft that it could not be removed without picking away
the gum from the proximal face. The ornament inserted in the gum
while still plastic consists of sections of the nassa shell, as in the two
pieces figured, of human molars, or of both in combination. This row
of ornament, which essentially consists of a series of plane faces, I
regard as carried over into the wooden metamorph in the slight saucer-
ing next the shaft in Plate V, figure 12. It has not been deemed nec-
essary to disarticulate any of the clubs of the type figured in this con-
nection, for I have had the opportunity to see the clubwright in the act

1 2 3 4 5 6 7 8 9 10 12 14 15 16 17 18 19 20 21

Online LibraryWilliam ChurchillClub types of nuclear Polynesia → online text (page 12 of 21)