View
234
Download
2
Category
Tags:
Preview:
DESCRIPTION
Extensive research on cone shells found in the P.N.G area and tracing theit carved motifs to some of the earliest expression of Oceanic Art.
Citation preview
Engraved prehistoric Conus shell artifacts from southeastern Papua New Guinea: their
antiquity, motifs and distribution
WAL AMBROSE, FIONA PETCHEY, PAMELA SWADLING, HARRY BERAN, LIZ BONSHEK,
KATHERINE SZABO, SIMON BICKLER and GLENN SUMMERHAYES
ABSTRACT
From c 1000-500 BP distinctive, engraved shell valuables made from large Conus are part of the material
culture found along the northern part of the eastern tip of New Guinea and on islands in the northern Massim.
They are a signature artifact of the Expansion Phase, a period of growth marked by an increase and expansion
of archaeological sites in this region (Egloff 1972; 1979).
The artifacts, with one exception, are decorated with engraved curvilinear decoration making them the oldest
radiocarbon dated artifacts decorated in the Massim art style. Undamaged examples consist of complete shells
modified in such a way as to allow hanging on a rope, but use as pendants would have been unwieldy and the
hole in each artifact is too small to allow wear as armshells.
By c 500 BP contact between New Guinea and the northern Massim has declined and the islands of the northern
Massim are part of a specialised insular trading system. A very different Conus valuable becomes one of two
shell artifacts (viz. an armshell and a necklace) that when exchanged for the other demarcates the extent of the
kula circuit.
Keywords: Conus shell valuables, exchange, kula, late Holocene, oceanic reservoir effect, Milne Bay Province,
Massim art, Oro Province, Papua New Guinea, sampling museum artifacts, trade
Introduction
by Pamela Swadling and Harry Beran
The engraved Conus artifacts from southeastern Papua New Guinea have long been recognised for their artistic
merit (e.g. Monckton 1905; Pöch 1907a; Seligman and Joyce 1907; Specht 1988 and White et al. 1970), but
their age has remained uncertain. Thirteen were found in prehistoric midden mounds at Rainu in the Wanigela
area (Collingwood Bay, Oro Province) in the first decade of the 20th century and since 1971 another 19 have
been reported from widely separated islands in the Massim region, which coincides approximately with Milne
Bay Province. These Conus artifacts are of particular interest as they are the earliest radiocarbon dated artifacts
known decorated in the Massim art style, one of the major art regions of Papua New Guinea.
The sampling and restoration of three Conus shell artifacts (one from Rainu in the Wanigela area and two from
Budibudi Atoll) by Wal Ambrose and dating by Fiona Petchey provide the foundation from which the rest of
this paper developed. As explained below, the one from Wanigela dates to 600-455 calBP / AD 1350-1495 and
WA: Archaeology and Natural History, School of Culture, History and Language, Australian National
University, wal.ambrose@anu.edu.au; FP: Radiocarbon Dating Laboratory, Waikato University, NZ,
fpetchey@waikato.ac.nz; PS: Archaeology and Natural History, School of Culture, History and Language,
Australian National University, pamela.swadling@anu.edu.au; HB: hberan@btinternet.com; LB: Donald Horne
Institute for Cultural Heritage, University of Canberra, elizabeth.bonshek@canberra.edu.au; KS: University of
Wollongong, Australia, kat.szabo@gmail.com; SB: Bickler Consultants Ltd, Auckland, NZ, arch@bickler.co.nz
; GS: Anthropology, University of Otago, NZ, glenn.summerhayes@otago.ac.nz
2
those from Budibudi Atoll 850-650 calBP / AD 1100-1300. The other contributors examine the nature of the
artifacts and their regional differences, where and how they were found, their iconography and how they fit in to
what is known of the prehistory of eastern Papua New Guinea.
Figure 1: Location map showing the offshore islands and the eastern tip of New Guinea
The known finds of engraved Conus artifacts
by Harry Beran, Simon Bickler and Pamela Swadling
All the engraved Conus artifacts found at Rainu in the Wanigela area were unearthed and this also seems the
case in the Massim region. Little is known about their archaeological context in either the midden mounds at
Rainu or within the Massim. With the exception of two finds from Lidau on Woodlark Island that were exposed
by a bulldozer on a logging road, the other Massim finds seem to have been found in coastal locations. The
Fergusson Island find was collected in an inland village, but the presence of marine worm activity on the inside
of its shell indicates that the artifact was found in a coastal area. Katherine Szabo describes the physical
condition of the two recently dated Budibudi finds. One of these shows that subsequent to its manufacture and
initial use it was exposed to a marine environment. After its discovery in more recent times some surfaces on
this Budibudi artifact were reworked.
Thirty-two Conus artifacts are held in public and private collections and all but six are illustrated below. Figure
2 gives the terminology used to describe Conus shells in this paper. Table 1 is a list of the eleven shells that
have been published and by whom.
Table 1: List of the eleven engraved Conus artifacts that have been published and by whom
3
Cat. No. Publication
E12961 Specht (1988: 19)
E15596A Bellwood (1979, Fig. 9.29), Specht (1988: 19)
E15597 White et al. (1970), Tets (1971), Specht (1988: 19), Meyer (1995: 154)
E15598 Specht (1988: 19)
Oc1905,0209.336 Seligman and Joyce (1907, Plate VIII)
Oc1905,0209.337 Monckton (1905, 1922: opp. 116) and Seligman and Joyce (1907, Plate VIII);
Oc1905,0209.338 Monckton (1905, 1922: opp. 116), Seligman and Joyce (1907, Plate VIII), and Newton (1988, Fig. 17, part view)
78712 Pöch (1907a, Abb. 7C and 8);
Imdeduya Malnic with Kasaipwalova (1998: 16)
Nuratu Mackay (1971), Leach and Leach (1983, Plate 6), and Malnic with Kasaipwalova (1998: 107)
Bickler 5 Bickler (2006:49)
Table 2: Wanigela finds
Find location Collector Current location Cat. No./ name height
cm
width
cm
1 Rainu, Wanigela Percy J. Money Australian Museum, Sydney E12961 9.97
2 Rainu, Wanigela Percy J. Money Australian Museum, Sydney E15595 7.94
3 Rainu, Wanigela Percy J. Money Australian Museum, Sydney E15596A 8.63
4 Rainu, Wanigela Percy J. Money Australian Museum, Sydney E15596B 9.36
5 Rainu, Wanigela Percy J. Money Australian Museum, Sydney E15597 10.65
6 Rainu, Wanigela Percy J. Money Australian Museum, Sydney E15598 11.56
7 Rainu, Wanigela C.A. Monckton British Museum, London Oc1905,0209.336 10.0 ??
8 Rainu, Wanigela C.A. Monckton British Museum, London Oc1905,0209.337 8.5
9 Rainu, Wanigela C.A. Monckton British Museum, London Oc1905,0209.338 8.5
10 Murin, Wanigela Rev. A.K. Chignell British Museum, London Oc1912,+.158 12.0
11 Rainu, Wanigela Rev. A.K. Chignell Horniman Museum, London 9-75 11.0
12 Rainu, Wanigela C.J. Seligman Cambridge Museum, England 1923.80 10.0
13 Rainu, Wanigela R. Pöch Museum für Völkerkunde, Vienna 78172 9.5
Notes: Murin is more likely to be where a villager who found it at Rainu lived than the find spot of Oc1912,+.158
With the exception of Oc1905,209,336 width measurements are not given as the shoulder region of each artifact has been removed
Table 3: Massim finds
Find location Collector Current location Cat. No./ name height
cm
width
cm
1 Wawela, Kiriwina Is J. Kasaipwalova unknown Imdeduya
2 Nuratu Islet, Kitava Is - in kula since 1971 Nuratu 17.5
3 Lidau, Woodlark Is private collection Milne Bay Natasha
4 Lidau, Woodlark Is private collection Milne Bay Nicola
5 Budibudi Atoll R. Aldridge Jolika collection, de Young Museum JFB.088.1 16.5 10.3
6 Budibudi Atoll R. Aldridge Jolika collection, de Young Museum JFB.088.2 15.8 8.9
7 Budibudi Atoll R. Aldridge Jolika collection, de Young Museum JFB.0141 / Bickler 1 15.0 10.5
8 Budibudi Atoll - in the kula Bickler 2 17.6 12.4
9 Budibudi Atoll - in the kula Bickler 3 17.5 10.1
10 Budibudi Atoll - in the kula Bickler 4 17.0 10.2
11 Budibudi Atoll - in the kula Bickler 5 17.5+ 11.3
12 Fergusson Is R. Aldridge Harry Beran collection HB II/53 14.5 9.4
13 Wabulak Is, Budibudi Atoll John Heaton private collection Heaton 1 ?? ??
14 Wabulak Is, Budibudi Atoll John Heaton private collection Heaton 2 12.7
15 Wabulak Is, Budibudi Atoll John Heaton private collection Heaton 3 15.5
16 Wabulak Is, Budibudi Atoll John Heaton private collection Heaton 4
17 Wabulak Is, Budibudi Atoll John Heaton private collection Heaton 5
18 Wabulak Is, Budibudi Atoll John Heaton private collection Heaton 6
19 Wabulak Is, Budibudi Atoll John Heaton private collection Heaton 7
4
Figure 2: Terminology used in describing Conus shells
The Wanigela context and finds
by Elizabeth Bonshek and Pamela Swadling
The engraved shells were found when the Wanigela mission station was relocated from an inland location to
Rainu on the coast of Collingwood Bay. The first mission station had been established in 1898 adjacent to a
fortified inland village called Wanigela. At the time of the missionaries’ arrival the fortified village contained
an alliance of clans whose origins stretched out along Collingwood Bay. In 1904 when the mission relocated to
the unoccupied coastal strip the villagers at Wanigela decided to move with the missionaries. They established
two villages, namely Rainu and Oreresan on either side of the mission station. Since then the name Wanigela
refers to the area rather than a specific settlement (Egloff 1979:1, 19). In the 1900s the Wanigela area was
inhabited by a number of groups connected to one another by partnerships that entangled them in networks of
assistance in raiding, trade, exchange and intermarriage (Egloff 1979; Bonshek ?).
The thirteen engraved shells were found when ancient midden mounds were levelled to infill swampy ground
during the construction of the Rainu site. In addition to being levelled, one of these mounds (that described by
Money 1905) has since been eroded by the sea. A survey of the proposed mission site at Rainu made in 1903
indicates that at the time the shoreline was 30 metres further out (Egloff 1979:10). Indeed today erosion
continues to be a problem and many villagers have moved their houses to sites further inland.
Two Anglican missionaries, one government officer and an Austrian ethnologist were responsible for collecting
the engraved shells now held in museum collections (Table 2). Percy Money (a layman) and the Rev. A.K.
Chignell were based at the mission at this time. C.A.W. Monckton was the Resident Magistrate who with
Money’s assistance had established the government station at Tufi in 1901 and subsequently surveyed the
mission site at Rainu. Monckton collected three shells and placed these in the British Museum. Percy Money
collected six finds and these are in the Australian Museum in Sydney. The Rev. A.K. Chignell collected two
and they are held respectively at the British and Horniman museums in London. Rudolf Pöch reports finding
one engraved shell in his excavations, it is now in the Museum für Völkerkunde in Vienna. How the
anthropologist Charles G. Seligman obtained the shell he deposited at the Museum of Archaeology and
Anthropology in Cambridge, England is not recorded in their catalogue. Seligman was in correspondence with
C.A.W. Monckton and did send him chemicals that were used in an unsuccessful attempt to conserve skeletal
5
material found in the mounds; in return it is possible that Monckton sent him a shell (Chignell 1911; Egloff
1971; Gash 1975; Monckton 1905, 1922; Pöch 1907a, 1907b and 1907c; Seligman and Joyce 1907).
The mounds to the south and west of Oreresan village became the focus of Brian Egloff’s 1967 archaeological
investigations, as the mounds in the Rainu area had been levelled. Egloff was unable to investigate the largest
surviving mound (his mound A) as it had been a cemetery for 60 years, but was able to excavate three mounds
he calls B and D above the Sasap River, as well as C located to the west of Oreresan village. A number of
radiocarbon dates were obtained for these three mounds. They date from 1410 calBP to 680 calBP (Egloff
1979:19-20, 24-25, 29, 31; Brian Egloff pers. comm. 2011).
The excavations posed challenges with respect to discerning the stratigraphy of the middens. Even though the
excavations were extensive and yielded large quantities of potsherds they seldom revealed clear-cut
stratigraphic layers. The basal deposits of mound C have the best representation of the earliest pottery
documented for the Cape Nelson-D’Entrecasteaux Island region, but the bulk of the mound deposits consist of
pottery made during the Expansion Phase that dates from c 1000-500 BP / c AD 950-1450. We now know that
this is the phase when the engraved Conus shell artifacts were made and used, but Egloff did not find any
examples in the three mounds he excavated. The upper deposits of mounds B and D have pottery assigned by
Egloff to the Refuge Phase starting c 500 BP, as the other sites that have the same pottery are in marginal areas,
such as the extensive swamps at the mouth of the Anina River (Egloff 1979: 20, 85-87, Brian Egloff pers.
comm. 2011).
Over the past century there has been a decrease in the range of traditional material culture items made in
Wanigela. Formerly men would have carved paddles, spears, shields, clubs, headrests and lime spatulas,
architectural features and canoes from wood. They also made body ornaments (some continue to do so)
originally worn in warfare, hunting and dancing, but which today are only worn when dancing. Carved wooden
items now housed in museum collections reflect a preponderance of rectilinear designs, but objects associated
with material culture made by women (for example designs on pots) reflect a visual language that includes
curvilinear motifs.
Although the engraved Conus artifacts found at Rainu are predominantly decorated with curvilinear designs,
one artifact (1905,209.336) is atypical. Its unmodified shell and rectilinear rather than curvilinear designs make
it stand out. The motifs consist of chevrons forming either a diamond or an X, depending on placement. They
closely resemble a motif used in historic times to indicate chiefly status. Today this motif is found on some lime
spatulas in museum collections, it also appears on some pot rims and now graces the Church of St Peter at Sarad
Mission near the Wanigela airstrip.
The Massim context and finds
By Harry Beran, Simon Bickler and Pamela Swadling
Nineteen engraved Conus shells have been found in the Massim islands, the great majority at Budibudi Atoll.
Finds from Budibudi Atoll
Fourteen engraved shells have now been collected from Budibudi Atoll with at least seven of the known finds
originating from Wabulak Island. When Simon Bickler did his archaeological fieldwork in 1995-6 five
engraved shells from Budibudi were held in a southeast Woodlark village. On the basis of the brief field survey
he did at Budibudi Atoll Bickler suspects that the engraved shells he saw on Woodlark had been found on
Wabulak Island. The offcut produced when one of the engraved shells was made into a kula armshell on
Woodlark was used to obtain the first radiocarbon date for an engraved Conus artifact from the Massim (Bickler
6
2006). Another shell (Bickler 5) was cut after Bickler’s departure and collected as a kula valuable on Budibudi
by Richard Aldridge along with two other engraved Conus now held in the Jolika collection of the de Young
Museum. The following report on these two Conus by Szabo shows that one of these artifacts has spent some
time in a marine environment.
In August 1997 four passengers on board a charter boat obtained seven engraved shells from Budibudi Atoll.
They recorded the following information as to how they had been found. An old man on Wabulak Island:
explained how a few months before he and his companions went out to gather coral rocks at low tide in
view of repairing the foundation of a communal hut. That is when they stumbled upon a coral rock
mound from which they decided to gather the stones. When excavating the mound about half way
they found the shells. They were bundled together as white birds in a nest. The old man thought they
had been placed there a long time ago and as long as he knew they had been protected from the elements
within this cocoon of stones. It sounded as if the structure he described had been purposefully designed
as a cache to hide and protect the shells for decades or centuries (John Heaton pers. comm. 2010).
The old man possessed six of the engraved shells while another was held by a man on Budalun Island who had
obtained it not long before on Wabulak Island. The Budalun Islander confirmed the account as to how the
engraved Conus shells had been found on Wabulak Island.
Imdeduya found at Wawela, Kiriwina
John Kasaipwalova (pers. comm. 2009) has told Harry Beran that he called this shell Imdeduya after a beautiful
woman of this name in a kula epic entitled Yolina. He brought it from a relative at Wawela who had found it at
the old village site of Wawela. In about 1976-79 the artifact had been washed out of the lower strata of a sand
dune by the sea. Kasaipwalova added:
I had [the shell] in my possession for . . . twenty years but when my father was sick and towards his final
year I presented it to him as a farewell ‘headrest’. On the day he passed away I was not present and in
the course of ‘caretaking’ of valuables at his death this was taken. I do not know who has it now or
whether it has been stolen.
For accounts of the Yolina myth see Leach (1981) and Senft (2008).
Nuratu, found at a small uninhabited islet of this name off the west coast of Kitava
This engraved shell was already in the kula in 1971 (Mackay 1971) and was still in it in 2009 (John
Kasaipwalova, pers. comm. 2009). Note that the bead decoration was changed between the report by Mackay
(1971) and the subsequent photograph in Leach and Leach (1983: Plate 6).
Natasha and Nicola found at Lidau, Woodlark Island
These two carved shells were found at Lidau on Woodlark Island in the late 20th century by a brother of
Pelupelu of Kitava, or that brother’s wife, while making gardens after bulldozers had made a logging road
through to Lidau. Pelupelu obtained them from his brother and gave them to John Kasaipwalova. The latter has
told one of the authors that ‘when Pelupelu brought the two [shells] to me, he had already decorated them and
even had them "baptised" as Natasha and Nicola by indelible ink . . . for my twin daughters Natasha and
Nicola’. In the 1990s, Kasaipwalova ‘placed’ the shells with John Kanadi hoping to raise school fees for his
children. When Kasaipwalova saw Kanadi in 2009, the latter said that the shells were with his wife Judy, from
whom he had by then separated, in Anagusa (Bentley) Island in the Engineer Group (John Kasaipwalova, pers.
comm. 2009).
Shell found on Fergusson Island
This engraved shell was collected in an inland village of Fergusson Island. Swadling has observed evidence of
marine excavating sponge (Cliona sp.) activity on the inside of the shell (see also report below by Katherine
Szabo on the presence of this sponge on one of the Budibudi finds). This activity indicates that after being made
into an artifact the shell had been buried near the seashore. When discovered in more recent times it was
7
reworked to make a kula valuable by drilling holes along the lip so that decorations can be attached.
The physical state of two Conus artifacts from Budibudi (JFB.088.1 AND JFB.088.2)
by Katherine Szabo
These two large, carved Conus cf. leopardus shells have a number of surface features that reveal particular
aspects of their history as artifacts. Both show evidence of reworking, while the larger example shows evidence
of discard or loss followed by recovery and recirculation as an artifact. The two artifacts are discussed in turn.
Conus artifact JFB.088.1
This artifact is considerably more worn and damaged than the smaller specimen. Differential wear and damage
on the various worked surfaces give clues to its life history as an artifact. The most revealing modification to
the shell’s surface are the numerous small pits created by the excavating sponge Cliona sp. (Porifera:
Clionidae). An important ‘bioeroder’, clionids attack carbonate surfaces in shallow water marine environments
producing rounded tunnels and galleries in the substrate (Calcinai, Arillo, Cerrano and Bavestrello 2003).
Clionid borings are frequently found on marine shells, and are sometimes even found actively excavating into
the surfaces of shells of living molluscs (Vermeij 1978:64). However, if a mollusc still occupies the shell, pits
will only be found on the outer surface of the shell. If clionid pits are located on the inner shell surface, or on
fracture surfaces of broken shells, then the damage certainly happened after the death and decay of the mollusc
(see Walker 1998:111).
For the purposes of shell artifact analysis, clionid borings on worked surfaces (excluding the shell exterior)
clearly indicate that the artifact was redeposited back into seawater after working. Such is the case with artifact
JFB.088.1. There is clionid damage on ground and chipped surfaces – primarily, but not exclusively, on the
ventral surface (aperature side) of the shell. However it is clear that there has also been an episode of grinding
in the anterior portion of the shell after the occurrence of clionid damage. This is demonstrated by the
sectioning of a number of clionid pits during the grinding process, exposing inner chambers and walls of clionid
excavations. The action of bioeroders thus tells us that the shell was initially worked, deposited back into the
sea, and then recovered and reworked.
Despite the extensive damage caused by clionids, it is difficult to estimate the minimum amount of time the
artifact may have been underwater before retrieval. Experiments investigating bioerosion rates and CaCO3 loss
in various natural materials revealed a loss rate of 17.59 kg m2 (dry weight) in Tridacna squamosa on the Great
Barrier Reef (Schönberg 2002:317). However the specifics of depositional environment and other
environmental variables that influence growth rates and reproduction make firm estimates of shell loss over
time difficult (Rützler 1975:215). The Conus artifact discussed here, could have remained underwater for a
period of months to years.
Examination of the worked surfaces provides further evidence for ongoing modification of the artifact. Apart
from the sequence of grinding and bioerosion damage, other surfaces display differential wear. The worked
edge running parallel to the spire is heavily rounded as is the spire itself where the apex has been removed to
produce a hole. This contrasts with the reduced lip of the aperture, which shows a mixture of older and fresher
breaks.
Conus artifact JFB.088.2
The traces of working on the smaller of the two specimens closely match those seen in the larger example,
including grinding along the split body whorl of the shell, a rounded and heavily worn break running parallel
the spire sutures, and an aperture lip region showing evidence of more recent breakage. The major
differentiating feature is that there is no clionid damage to indicate a period of loss/discard of the artifact.
8
The description of the prehistoric material
by Pamela Swadling, Harry Beran and Simon Bickler
The shells from the Massim and Wanigela have different shapes; see Types 1 and 2 in Figure 3. Those from
Wanigela were originally Type 1 artifacts that were cannibalised for their armshells in the prehistoric past. The
effort put into shaping the Conus shell artifacts as well as the engraved decoration on these artifacts indicate that
they were items of value.
The Massim artifacts (Type 1) have a hole in the spire cap, the inner whorls have been removed and the upper
part of the body whorl has been cut at right angles to the aperature. Twelve of the 13 examples for which we
have photographs are worked in this way. The Wawela (Kiriwina) shell has a small hole in the spire cap, but it
is not clear from the available illustration whether the body whorl has been cut. It certainly lacks the gouged
notch on the upper body whorl of the shell.
The small size of the hole in the spire cap of the Massim artifacts makes them useless as armshells and they
would be cumbersome as pendants, but their careful shaping and decoration indicates that they were valued
artifacts. The hole through the spiral cap and right angle cut into the body whorl would have allowed a rope to
be threaded through so that multiple artifacts could be strung together. Perhaps in a comparable way to the 22
armshells shown in Malinowski’s photo taken on Kiriwina in 1915 (Young 1998:229, plate 155). These mwali
were strung and hung from a pole so that two men could carry them. When not strung on a rope the Type 1
artifact can stand independently if it is placed on its spire cap end. Most of the motifs are best viewed in this
way. This is the orientation used in the illustrations provided in this paper. It may be the intended display
orientation. When placed this way the bird on E15597 from Wanigela is aligned with the land or sea whereas
anthropomorphic faces are orientated to the sky (see Oc1905,0209.338 from Wanigela).
9
Figure 3: The difference between most of the prehistoric Conus artifact finds from
the Massim and those from Wanigela
Most of the Wanigela Conus artifacts are arc-shaped pieces of shell. The spire cap, inside whorls and upper part
of the body whorl are absent. Only 2 of the 13 known finds from Wanigela are not arc-shaped. Most Wanigela
artifacts were originally Type 1 artifacts from which the armshell section has been removed. In some cases the
armshell was cut off at the notch leaving a bevelled edge, as in E12961. Some shells have a regular edge, such
as Oc1905,0209.337, whereas others are more jagged and irregular such as Oc1905,0209.338. On a number of
shells, the cuts were made without any attempt to preserve motifs on them intact, for example, on
Oc1905,0209.337.
Two investigators, Monckton (1905: 33) and Pöch (1907a: 68), report an association between engraved Conus
shells and human remains. Monckton writes that in excavations that had been carried out by ‘the mission and
natives of Rainu’, two fragmentary engraved shells (namely, Oc1905,0209.337 and Oc1905,0209.338) had been
found among human remains 4 feet (120 cm) below the surface; he adds, ‘placed, I think, originally in graves’.
In his own excavations at Rainu Pöch found an engraved shell (namely 78172), perhaps a piece of jewellery,
among four human skeletons, approximately 1 metre below the surface. He goes on to say that the mound he
had excavated was probably a midden (Abfallshügel) on which a house had been built and that the deceased had
been buried under this house. He also found pig bones, potsherds, and shell knives in the same midden. Pöch’s
remarks leave it unclear whether the engraved shell had been buried with the deceased or thrown away with
10
animal bones, potsherds, and shell knives. Nevertheless, Monckton’s and Pöch’s remarks point to the possibility
that the engraved shell off-cuts, created when the artifacts were cannibalised to make armshells, may have been
valued enough to have been used as grave goods.
No wide armshells, as used in the kula, have been excavated from the Wanigela middens. The only armshells
found are a number of narrow ones made from Conus and Trochus shells (Seligman and Joyce (1907, Fig. 4;
Egloff 1979, Plate 13).
Apart from extracting armshells some of the remaining offcuts may have been used to make longitudinal shaped
small adze blades. One artifact, E15595, is a reduced arc-shaped piece of body whorl with an unfinished
longitudinal cut that goes through engraved designs. It is possible that this artifact was being cannibalised for
raw material to make small Conus shell adzes. Extracting shell in this way may also explain the reduced arc of
some artifacts such as E15598. Malinowski (Young 1988:265) reports that the Mailu once made these artifacts,
but at the time of his visit in 1914 they were no longer being made nor were discarded artifacts available for
him to purchase.
There are two unusual finds from Rainu. The apex of E15596-B has been cut away, but the inside body whorl
has not been trimmed back to produce an arc-shape piece of shell. The second exception (Oc1905.209.336) is
very different from the other finds. This shell is complete and decorated in a rectilinear style with a series of
motifs each consisting of a diamond shape framed by nested chevrons. The motif is similar to that on the
supports of some contact-period headrests from Collingwood Bay, as illustrated in Grunne (1979, fig. 9.10),
Meyer (1995, fig. 157), Hamson and Aldridge (2009: 244-7), and Beran and Aguirre (2009, Ch. 6). Perhaps
someone trained to carve in a rectilinear style engraved the shell in Collingwood Bay, in imitation of the shells
with curvilinear motifs.
At Rainu engraved Conus artifacts were cannibalised in the prehistoric past to make armshells. They have also
been reused in the Massim. Since about 1971, a number of the engraved Conus shells unearthed there have gone
into the kula. Some have entered it as full-length shells (Nuratu and HBII/53 from Fergusson). In this case, the
only modification needed to make them into kula valuables was the drilling of a series of holes along the lip to
attach egg cowries and trade beads. Bickler 3, Bickler 4, and Bickler 5 appear also to have entered the kula at
full length as they are drawn by Joan Lawrence, based on Bickler’s photographs, with holes along the lip (see
Figure 8). Other shells have been cut to produce armshells like those normally used in the kula, called mwali
(JFB.0141 and Bickler 5). The shell JFB.0141 was photographed by Bickler as a full-length kula shell decorated
with egg cowries, but was later cut into a mwali-style armshell. The lips of the full-length shells JFB.088.1 and
JFB.088.2 are not drilled for attachments. As these lips are broken, the shells may not have been considered
suitable for kula.
Making prehistoric finds into kula valuables may relate to the scarcity of large cone shells as exemplified by the
splitting of paired mwali in order to increase the number of items available for exchange. During the early years
of colonialism a shortage of mwali arose due to changes that allowed more people to participate in kula
exchanges. To overcome this shortage paired armshells were separated and necklaces shortened (Liep and
Affleck 1983: 126).
11
Figure 4: Wanigela designs I
a. Cat. No. 9-75, height 11.0 cm
b. Cat. No. E15596-A, height 8.63 cm
c. Cat. No. Oc1905,0209,338, height 8.5 cm
d. Cat. No. 78172, height 9.5 cm; rolled out view on left is from Pöch 1907a:70, Fig 8
e. Cat. No. E15597, height 10.65 cm; dotted line above bird is line of artist’s correction
Unless indicated otherwise the drawings in Figures 4-9 are drawn by Swadling from photographs
12
Figure 5: Wanigela designs II
a. Cat. No. Oc1905,0209.337, height 8.5 cm
b. Cat. No. E15598, height 11.56 cm
c. Cat. No. E15595, height 7.94 cm
d. Cat. No. E12961, height 9.97 cm
e. Cat. No. Oc1912,-,158, height 12.0 cm
f. Cat. No. 1923.80, height 10.0 cm
g. Cat. No. E15596-B, height 9.36 cm
h. Cat. No. Oc1905,0209,336
13
Figure 6: Massim designs I: From Budibudi Atoll
a. Cat. No. JFB.088.1, height 16.5cm, AMS radiocarbon date 850-670 calBP
b. Cat. No. JFB.088.2, height 15.8 cm, AMS radiocarbon date 785-650 calBP
14
Figure 7: Massim designs II: From Wabulak Island, Budibudi Atoll
a. Cat. No. Heaton No. 2, height 12.7 cm
b. Cat. No. Heaton No. 3, height 15.5 cm
c. Cat. No. Heaton No. 1, height **
awaiting further photos
15
Figure 8: Massim designs III: From Budibudi Atoll
a. Cat. No. Bickler 1 subsequently cut to make a mwali, Cat. No. JFB.0141, see Figure 14.
b. Cat. No. Bickler 2
c. Cat. No. Bickler 3
d. Cat. No. Bickler 4
e. Cat. No. Bickler 5, AMS radiocarbon date calBP 882-632 calBP
Drawings of Bickler 1-5 are rolled out rather than natural representations. They were drawn for Simon Bickler
by Joan Lawrence.
16
Figure 9: Massim designs IV
a. Cat. No. HB II/53, inland Fergusson Island; height 14.5 cm
b. Cat No. Nuratu. Nuratu Islet, Kitava, Trobriand Islands; height 17.5 cm
c. Wawela, Kiriwina, height not known
Gaps in shell outlines indicate that this area was not visible in the available
photographs.
Unlike the others, the Wawela (Kiriwina) shell is decorated all over with thin incisions. The grid pattern incised
on the shell appears recent, possibly due to reworking, as the other incised designs appear to be contained
within these grids.
The iconography of the carved shells
by Harry Beran
It is potentially misleading to speak of a Massim (visual) art style because the designs on Massim woodcarvings
are executed almost entirely in a curvilinear style, while those on Massim pottery are applied in a predominantly
rectilinear style. Hence, one should speak of a Massim woodcarving style and a Massim pottery decoration
style. Judging by the quality and curvilinear style of the designs on the engraved shells, they were probably
executed by skilled woodcarvers. This observation suggests that men made the engraved shell artifacts.
No attempt has been made to consult Collingwood Bay or Massim informants on the meaning of the motifs on
the shells. Because of the age of the shells and uncertainty as to where exactly they were carved, it seems best to
compare the ancient motifs to those on contact-period art of which interpretations have been obtained from local
informants.
In both the artifacts from the Massim region and Wanigela there is variation in the depth of engraving or
incising; some shells have deep gouge-like incisions whereas others have thin line incisions. Moreover, the
17
motifs on some shells are merely incised while those on others are carved in low relief. The scale of the
drawings provided in this essay makes it impossible to indicate which are carved in relief, but Table 4 indicates
on which shells the motifs are incised and on which they are carved in low relief.
The prehistoric artists followed two design layout conventions as some shells have all-over decoration whereas
others have a frame design with a central void (compare shells Bickler 4 and Oc1905,029.338 in Figure 10).
The former practice is more common in the shells found in Wanigela, the latter in the shells found in the
Massim region.
On the Massim-found artifacts, motifs are placed in three main locations; namely, below the shoulder extending
in some cases to below the notch, along the sides of the aperture, and to the anterior of the shell. The
distribution has the effect of leaving a plain area in the centre of the shell. The drawing in Figure 10 shows this
design clearly. Only the shell found at Wawela (called Imdeduya) is decorated over most of its surface.
On the Wanigela-found shells, overall decoration is more common than frame decoration. However, shells
E12961, E15597, E15598 and Oc1912,+.158 are decorated like those found in the Massim region; they leave
the central area plain and have decoration fields below the shoulder, along the sides of the aperture, and at the
anterior of the shell.
Figure 10: Frame and over-all decoration
a. Cat. No. Bickler 4, Budibudi Atoll
b. Cat. No. Oc1905,0209.338, Wanigela; height 8.5 cm
Although the designs are rolled out in Bickler 4 and Oc1905,0209.338 has a natural representation it is clear
that the central area is void in the frame layout and infilled with motifs in over-all decoration.
Two shells are left aside in the following discussion because they are highly unusual. Oc1905,0209.336 is the
only shell engraved in a rectilinear style. As mentioned above, perhaps it is a copy of the shells carved in a
curvilinear style, executed in an area of Collingwood Bay where a rectilinear carving style predominated.
Imdeduya from Wawela is covered in interlocking scrolls but unlike the other shells with an all-over design, the
scrolls are arranged in three panels around the shell and incised very lightly. Also, judging from the only
photograph available, it is the only Massim-found shell where the internal whorls and much of the spire have
not been removed.
The motifs on the shells found in Wanigela and the Massim region are carved in a curvilinear style, as are
contact-period Massim woodcarvings, and there are many specific parallels between the motifs on the shells and
contact-period Massim art. Some of the motifs on the shells are carved in low relief and this too is a common
feature of contact-period Massim art.
The most common motif on the shells is that shown in Figure 11, a circle or scroll or spiral turning into straight
lines gradually diverging from each other. The motif runs parallel to the shoulder edge on many of the Massim-
found shells; for example, Nuratu from Kitava and the Fergusson find. This motif appears on both ends of
18
contact-period canoes, typically with a bird at the end of the diverging lines. The motif also appears on a wide
range of other contact-period carvings, such as lime spatulas and canoe paddles, presumably in an allusion to
canoe-ends. Figure 12 shows it on a canoe component. Massim informants invariably interpret the motif on the
left in Figure 12 as the head of a fish (with open jaws); for example, that of a long-tom, a fish with long jaws,
teeth, and a tongue. Of the bird, different interpretations are given; they include reef heron and pigeon (Beran
1988:34-5). Figure 13 shows a canoe-end, which has been radiocarbon, dated to AD 1440-1640, 95%
probability. The motif on it links the motif on the carved shells with that on contact-period artworks. Needless
to say, it cannot be assumed that the motif on the prehistoric objects also represented the head of a fish. The
‘fishhead’ motif also appears on two of the Wanigela-found shells, but not in the same position. On E12961 it
appears once and on E15598 at least twice. On the Massim-found shells, the animal-head motif is often reduced
to a mere notch (e.g. Bickler 3) and sometimes the motif looks like two animal heads facing each other with the
jaws merging in the middle (Bickler 2). The notch is deep and usually cuts through the shell to form one or two
slits. The motif consisting of two animal heads whose open jaws meet also occurs at the anterior of a number of
shells, as shown, for example, in the drawing of the Fergusson find. The slits may have served to attach leaf or
shell-disc tassels to the shells.
Figure 11. (on left). Animal-head motif on Nuratu. Cropped from illustration in Mackay (1971).
Figure 12. (on right). Carving from a canoe. After Haddon (1894, plate XII, nr 185).
Figure 13. Canoe-end collected by Harry Beran in Kwaiawata, Marshall Bennett Islands. Jolika Collection, New
York (HB 840). Photograph by John Bigelow Taylor, courtesy of John Friede. The image has been rotated
horizontally in order to show it in the same direction as the artifacts in Figures 11 and 12.
Bickler 1 in Figure 14 shows a motif that may speculatively be interpreted as a canoe, given the importance
canoes must have always have had in the Massim islands. This motif, in a variety of shapes, is still carved on
southern Massim platters in the contact period. The motif on the right (Figure 15) is from a bowl in the Jolika
Collection (HB 291).
19
Figure 14. (on left). The motif above the notch on the shell Bickler 1 may represent a canoe.
Figure 15. (on right). A canoe motif on a contact-period Massim bowl. Jolika Collection (HB 291).
The bird carved on E15597 (Figure 16) is the most beautiful of the motifs carved on the shells. Birds are still
among the most common representational motifs in contact-period Massim art. Tets (1971) interprets the bird
on the shell as a belonging to one of a number of species of cranes not found south of the equator. Although the
image on the shell is less stylised than most bird motifs in Massim art, it is not naturalistic enough to say what
bird it represents. However, it is perfectly consistent with the long-necked bird being a reef heron (or egret),
which does occur in the Massim region and is an important motif in the region’s art. One or two long-necked
birds always occur at the bottom of carvings attached to the wave-splitters of kula canoes called nagega.
Informants told Seligman (1909:33) on Murua (Woodlark Island) that the bird is a boi, which he translates as
reef heron. An example of a wavesplitter attachment with one long-necked bird at the bottom of the carved
section is shown in Figure 17. Shirley F. Campbell (2002:98) was told that the boi, which she translates as
egret, is an important motif on the washboards of kula canoes (called masawa) carved in Vakuta, the
southernmost of the Trobriand Islands. Tets (op. cit) draws attention to the curl carved at the bird’s tail on the
shell. Surprisingly this kind of curl also appears on some of the birds carved in the contact period; for example,
on the bird at the top of the carving in Figure 17.
20
Figure 16. Shell E15597 with a bird motif (photo copyright Australian Museum)
Figure 17. Wavesplitter attachment for a nagega kula canoe. British Museum Oc,+.3872. (Copyright: Trustees
of the British Museum.) [note if printing paper from pdf file- when Swadling did fig 17 was printed upside
down]
The shell shown in Figure 18 shows four instances of a motif Swadling interprets as a snail. The snail is a motif
in contact-period Massim art, but now it is represented as concentric circles (Narubutau 1975, fig. 1; Campbell
2002: 101, 107). Figure 19 shows how it is carved on a Kitava Island washboard. The significance of the motif
in contact-period Trobriand art is indicated by the following remark of John Kasaipwalova (Malnic with
Kasaipwalova 1998: 15, cf. 22-3):
The house of a snail carries the sign of a spiral. The spiral starts at a point, which the Trobriand Islanders
call the gum (rhyming with boom). From the gum the spiral flings outward and becomes larger and larger
until, in no time at all, it reaches the sky and beyond.
21
Figure 18. Snail motifs on Heaton 1. Drawing by Pamela Swadling
Figure 19. Section of a Kitava Island washboard showing a series of snail motifs. Cropped from photograph by
John Bigelow Taylor, courtesy of John Friede. Collected by Ralph Lawton in Kitava Island. Jolika Collection,
New York (HB II/16).
Seligman and Joyce (1907:332) interpret the motif in the centre at the (broken) shoulder edge of
Oc1905,0209.338 as a face (Figure 20). If it is indeed a face then the most prominent motifs on E15596A and 9-
75 may also be faces, though even more stylised. Face-like motifs do occur in contact-period Massim art, for
example, on clapper lime spatulas where they apparently represent a lizard’s head (Beran 1988:27). There is no
reason to think that there is any connection between the face motif on the shells and face motifs on contact-
period Massim art.
Figure 20. The ?face motif on Oc1905,0209.338.
The same writers (op. cit.) comment that ‘the interspaces [between the interlocking scrolls on
Oc1905.0209.337] are filled with a sort of hatched leaf pattern’. It is clearest in the bottom right-hand corner of
the first drawing from the right of this shell in Figure 5. I am not aware of such a motif in contact-period
Massim art.
All the shells, except Oc1905,0209.336 engraved in a rectilinear style, display scrolls and spirals. The scrolls
sometimes form rows of interlocking scrolls (e. g. JFB.088.1 and Bickler 2). The spirals are sometimes
composed of two interlocking spirals (e.g. 78172). Similar interlocking scrolls are a common feature of contact-
period Massim woodcarvings (e.g. Newton 1975), but spirals are rare in this art and I do not recall any instances
of interlocking spirals. A curved line with inward-turning scrolls appears on HB II/53 from Fergusson Island
and exactly the same motif appears on the underside of some bowls apparently from Goodenough Bay in the
mainland part of the Massim region (Figures 21 and 22). The particular example shown is from a bowl in the
John and Marcia Friede (Jolika) Collection (HB589).
22
Figure 21. Detail of HB II/53 from Fergusson 1sland Figure 22. Detail of bowl in the Jolika Collection
(HB 589)
The similarity of carving style and particular motifs on the shells and on contact-period Massim art leave no
doubt that the shells are carved in the Massim art style, are early examples of Massim art (cf. Pöch 1907a: 71
and Damon 2002: 114). Of course, scrolls and representations of things in the world also occur in the contact-
period art of other regions of New Guinea. However, it is significant that the Massim region is separated from
these by two neighbouring regions, Oro and Central provinces, whose art style is predominantly rectilinear.
Table 4. Summary of information on iconography
Shell
number
Incised or
low relief
Animal head at shoulder
or notch
Frame or all-over
decoration
Figurative
motif
Height cm
Fh full height
F fragment E12961 ?incised frame 9.97 (F)
E15595 incised all-over 7.94 (F)
E15596A incised all-over ?face 8.63 (F)
E15596B incised all-over 9.36 (F)
E15597 relief frame bird 10.65 (F)
E15598 incised animal heads on body of shell frame 11.56 (F)
Oc1905,0209.336 incised all-over, rectilinear 10.0 (Fh) Oc1905,0209.337 incised all-over 8.5 (F)
Oc1905,0209.338 incised all-over ?face 8.5 (F)
Oc1912,+.158 incised frame 12.0 (F)
9-75 incised all-over ?face 11.0 (F)
1923.80 incised ?design eroded 10.0 (F)
78172 relief all-over 9.5 (F)
Imdeduya incised all-over in panels unknown
Nuratu relief animal head frame c. 17.50
Natasha no image
Nicola no image
JFB.088.1 relief notch or animal head frame 16.5 (Fh)
JFB.088.2 relief notch & animal heads on
body of shell
frame 15.8 (Fh)
Bickler 1/ JFB.0141 ?relief notch frame canoe 15.0 (Fh)
now ring
Heaton 1 relief notch frame snails
23
Heaton 2 notch frame 12.7 (Fh)
Heaton 3 XX XX 15.5 (Fh)
Heaton 4 no image
Heaton 5 no image
Heaton 6 no image
Heaton 7 no image
Bickler 2 ? animals head frame 17.6 (Fh)
Bickler 3 ? notch frame 17.5 (Fh)
Bickler 4 ? notch frame 17.0 (Fh)
Bickler 5 ? notch frame 17.5 (Fh)
HB II/53 Fergusson
1s
relief animal head frame 14.2 (Fh)
Radiocarbon dating of the engraved shells: sampling, dating, and the significance of the results
by Pamela Swadling
Four engraved Conus artifacts have now been radiocarbon dated. Bickler (2006:48-49) reports a date of 882-
632 calBP for an engraved shell, which was collected on Woodlark but originated from Budibudi Atoll some 70
km to the southeast. This date was obtained from the offcut produced when making the artifact into a kula
armshell (mwali). Three more AMS (Accelerator Mass Spectrometry) radiocarbon dates for engraved Conus are
reported here. Two were collected on Budibudi Atoll and the other from Rainu. John Friede made the two
Budibudi shells available and the de Young Museum funded their dating. The Australian Museum made
E15596-B available and Edward Aguirre funded its dating. This project would not have been undertaken
without Wal Ambrose’s skill in sampling and restoring these artifacts to their original appearance.
Out of the twelve Conus shells with curvilinear engraved motifs found at Rainu, E15596B was chosen for
dating as alone among them it retains its spire column. This structural feature made it possible to sample the
shell without disfiguring the artwork on the outer surface of the artifact. This is fortunate, as none of the other
shells could be sampled without fear of disfiguring their artwork.
Fiona Petchey of the Waikato radiocarbon laboratory oversaw the dating of the shell samples and also arranged
for the dating of two shells collected prior to atomic testing to determine the oceanic reservoir effect in Milne
Bay Province. This information gives the correction required for shell dates from this region. These results are
reported in detail below. The two shells JFB.088.1 and JFB.088.2 gave dates of 850-670 and 785-650 calBP
respectively, very comparable to the result obtained by Bickler (2006) for Bickler 5. The Rainu shell, E15596-
B, gave a date of c 600-455 calBP. The four artifacts are illustrated in Figures 5, 6 and 8.
All four artifacts come from the Expansion Phase dating from c 1000-500 BP/ c AD 950–1450, a period when
there were cultural links between the people of Collingwood Bay on the mainland of New Guinea and the
islands of the northern Massim, as far east as Budibudi Atoll.
The three Budibudi finds all come from the middle of the Expansion Phase whereas the Rainu artifact dates
from the end of this period. The latter date is extremely useful as it suggests that there may have been a stylistic
change during the period of their production. The decoration on the three Type 1 (Figure 3) Budibudi artifacts is
frame-like in style, whereas the Rainu artifact would have been a complete shell when its overall decoration was
engraved. This suggests that complete shells and overall decoration were more in vogue towards the end of the
Expansion Phase. The only other engraved complete shells are oddities, perhaps because they were produced
during a time of social change. These artifacts are the one with rectilinear designs from Rainu and the one with
curvilinear designs within a rectangular grid pattern from Wawela on Kiriwina.
Given that engraved Conus artifacts were being produced towards the end of the Expansion Phase it is likely
that the engraved offcuts found in Rainu mounds date from the Refuge Phase, that is post c 500 BP. They can
24
best be seen as heirloom artifacts that have been cannibalised for their potential armshells and are probably
grave goods associated with burials in the mounds.
Shell sampling for AMS dating and the restoration of three
Conus artifacts: two from Budibudi Atoll and one from Wanigela
by Wal Ambrose
All three artifacts sampled are museum specimens. They were only made available on the understanding that
the artwork would not be impacted and that they would be restored to their original appearance. It was possible
to sample the Budibudi artifacts by extracting shell from the interior of the shells where the jagged edges of
remnant, inner whorls provided sufficient material for radiocarbon dating. Restoring the extracted material with
a replica required moulds to be made of the affected area. A similar procedure necessitating multiple silicone
impressions was required for sampling and then restoring the remnant, exposed spiral stem of the Wanigela
artifact.
Budibudi Atoll
Two Conus shells from Budibudi Atoll with decorated surfaces were presented for sample extraction. The
surface of JFB.088.1 was pitted with a marine weathered appearance. In contrast the surface of JFB.088.2
seemed intact. After discussion with Pamela Swadling it was decided to extract the shell sample from the jagged
edges of the remnant inner whorls that had been removed in the preparation and manufacture of the engraved
artifacts, Figure 23. No decorated surface would be affected by this procedure. The edges removed for dating
would need to be replicated in order to restore the appearance of both artifacts.
Procedure
1 A silicone mould was made of the broken edges of the inner core whorl of each artifact. This first
impression would form one half of a two-piece mould needed for the restoration in resin of the sampled
shell whorl fragments.
2 A thin diamond dental blade, 25 mm in diameter, and mounted on a flexible coupling was used to dry cut
the remnant broken whorl edges of each shell. The cutting produced more shell in powder form than a
previous experimental extraction from a modern shell. This seemed to indicate that weathering had
reduced the proportion of collagen in the two artifact shells compared with the modern specimen. About
10 g of mixed shell solid and powder was removed from each Conus artifact.
3 A second silicone mould was made of the diamond cut surface of the inner whorl from where the broken
shell fragments were removed for radiocarbon dating.
4 A third silicone mould was made of the second silicone mould surface bearing the impression of the cut
surface. This third mould presented the same cut surface of the sampled shell and was used as the second
half of the two-piece mould mentioned above in procedure 1.
5 Each half of the two-piece mould section was then matched and marked to a correct register in a circular
collar from a short section of PVC pipe.
6 Epoxy resin, coloured to approximate the original shell, was run into the first silicone mould. The second
silicone mould was then pressed into the epoxy to form a cast matching both the original jagged surface
and the exposed cut surface.
7 The cast epoxy piece for each shell was fitted to the vacant sampled section. Some undercut and
protruding fractions of each epoxy cast had to be removed by grinding in order to fit the curved sections of
the shell. A polyvinyl butyral - ethanol cement was used to attach each cast piece to its appropriate shell
surface. The cast pieces could be removed if required with ethanol as a solvent.
25
Figure 23. Conus JFB.088.2 showing the epoxy cast of the
jagged whorl coloured to approximate the original shell.
Wanigela
In sampling this artifact (E15596-B) it was critical that neither the engraved surface nor the very delicate
remnant spiral whorl surrounding the inner spiral stem were damaged. The most accessible section was the
shell’s remnant protruding spiral stem. In discussion with Pamela Swadling it was decided that the latter should
provide the sample, including its delicate spiral septum (Figure 24).
The restoration of the removed section required a facsimile to be reattached in the same position as the removed
sample. The small protruding stem section of the inner whorl represented the fragile early stage in the shell’s
growth. In order to produce a replacement cast of the targeted section a thin flexible silicone mould was
required. A conventional two-piece mould would require masking one half of the stem with a wax ‘platform’
that would require working around the delicate remnant septum. This strategy was not used.
Instead a thin (~.5 mm) one-piece silicone skin was made by applying three paint-like coatings to the stem. This
had the advantage that a skin that could be pealed from the shell without damaging the flange-like spiral septum
and also made it easier to extract the polymer replacement part (Figure 25). The resulting silicone impression
would have the form of a rubber glove finger but have an inner surface that records a precise impression of the
original spiral stem and septum. A rigid outer casing was needed to maintain the thin silicone skin in its correct
original shape. This casing was formed in two stages using epoxy resin. Containment for the casting procedure
was made from a 20 mm diameter polycarbonate vial that was longitudinally sectioned to provide a trough into
which the resin could be added while the shell was supported horizontally (Figure 26). After the resin had set
both the half vial and the resin mould were removed, leaving the silicone skin still intact. The top of the resin
mould was then filed flat and several cupules were made in its surface to serve as registers for the second
matching epoxy mould. The first half of the epoxy mould was refitted into its original position supporting the
silicone skin. A second complete polycarbonate vial, reintroduced to the mould setting had a hole made to allow
for the injection of epoxy resin that would register with the lower first half to complete the rigid support for the
silicone skin. The final stage of the casting procedure was the removal of all the mould sections (Figure 27).
During this process a crack was formed across the shell’s fragile stem. It was the decided to use this as the
matching surface for the replacement facsimile. But before this could be done a second silicone impression of
the fracture surface of the remaining shell core was made. Eventually this minor mould would allow the
replacement shell stem, cast in epoxy resin, to be matched to the original crack surface.
26
After making coloured replacement components in epoxy resin (Figure 28) they were matched to the original
form of the removed shell and fixed together with epoxy resin having the same formulation and colour. A 1 mm
diameter hole was bored about 10 mm along the combined replacement to house a metal pin permanently fixed
in place with epoxy resin. In order to allow the replacement to be removed, if necessary, the replacement part
required adhesion to the Conus stem with a non-permanent adhesive. A ~1.5 mm diameter hole was bored into
the core to accommodate the removable metal pin already fixed permanently into the replacement part. The
larger diameter hole in the Conus shell core was to allow for any slight adjustment to allow matching between it
and the facsimile replacement. An ethanol solvent-based polyvinyl butyral resin was used to finally join the
newly made epoxy replacement spiral stem to the Conus shell artifact. Fine adjustment was made to match the
joining surfaces with the same solvent-based resin, coloured to match the original shell (Figure 29). This was
then left for several days to dry and form a stable bond. If necessary ethanol could be used to remove the newly
attached epoxy stem from the artifact.
A 1% polyvinyl butyral solution in ethanol was used to coat the shell stem as a resist before the silicone mould
was made. A diluted wax resist was used for separation of the epoxy sections and the silicone mould.
The final extracted shell stem weighed ~ 1 g. Although the minimum requirement for AMS dating was 80 mg it
was necessary to provide sufficient shell for laboratory cleaning and calcium-aragonite analysis to assess shell
recrystallisation.
Figure 24. A view of the protruding stem
selected for removal as the radiocarbon dating
specimen.
Figure 25. The stem has been encased in a
sheath of thin silicone rubber. A rigid
polythene baffle is seated between the shell
inner wall and the surface of the silicone
sheath. One-sided adhesive tape has been
affixed to the baffle and the shell wall. This
arrangement is designed to allow the epoxy
casing to be installed.
27
The dating of three Conus artifacts: two from Budibudi Atoll
Figure 26. The inner end of the horizontal
casing vial is seated into soft wax to form a
seal and allow liquid epoxy resin to be filled
into the half-vial.
Figure 27. The epoxy casing and inner silicone
rubber mould partly removed from the
enclosing vial to show the three components of
the casting system.
Figure 28. The original piece of the shell
stem selected for radiocarbon dating (top)
alongside the replica (below).
Figure 29. The reconstructed shell stem
28
and one from Wanigela
by Fiona Petchley
All samples were processed at the Waikato Radiocarbon Dating Laboratory. The shells were acid etched to
remove any post-depositional surface contamination and tested for recrystallization of the shell matrix.
Radiocarbon results are presented in Table 4 and Figure 30.
To obtain calendar ages for the two Conus sp. artifacts it was necessary to determine the local surface marine 14C reservoir correction factor, commonly called a !R. The marine !R is the difference between the global
average modeled marine reservoir - Rg(t) (e.g. Marine09: Rafter et al. 2009) and the actual 14C activity of the
surface ocean at a particular location –Rs(t) (Stuiver et al. 1986). This offset can be calculated from marine
samples collected from known locations prior to atmospheric bomb testing (~AD 1950), whose age of death is
known precisely (e.g. Petchey et al. 2004, 2008).
Ongoing research has indicated that wide shifts in !R are most common where ocean currents meet continental
landmasses, opposing currents, or island chains (Petchey et al. 2008). In the southwest tropical Pacific, waters
derived from the South Equatorial Current (SEC) flow around the southeast tip of Papua New Guinea and enter
the Bismarck Sea through Vitiaz Strait. Further north the northern section of the SEC is broken by the Solomon
Islands before entering the Bismarck Sea through the St Georges Channel between New Britain and New
Ireland. This dominant current flow is reversed during the winter (June/July) monsoon resulting in considerable
turbulence in the channels (Steinberg et al. 2006). Although a number of studies into marine reservoir variation
in the Bismarck Sea and Coral Sea region have been undertaken (Chappell and Polach 1976; Edwards et al.
1993; Petchey et al. 2004; McGregor et al. 2008) there is very little reliable !R data available, and it is likely
that considerable variability is possible in those locations where ocean currents are most turbulent. To monitor
possible variation within the Milne Bay Province and achieve as close a !R correction as possible for this
research we obtained two pre-AD 1950 shells from museum collections (Table 4). These came from Samarai
and Kiriwina (Figure 1).
Marine samples were calibrated using the Marine09 curve of Reimer et al. (2009) and a pooled !R of 40±15 14C yrs (!2
1:0.05 = 0.22<3.84) calculated from the pre-AD 1950 shells. All radiocarbon determinations were
calibrated using the OxCal program v4.10 (Bronk-Ramsey 2010).
29
Table 5. Radiocarbon data for shell samples from New Guinea
Lab. No. Location/
date of
collection
Material# ! 13C‰ 14C Age &
Error (BP)
[Rs(t)]
Calibrated Age (BP)
Wk-25782
(Cat. No.
JFB.088.2)
Budibudi
Atoll
Conus sp.
(C)
2.49±0.2 1209±30 739-670 (68.2%)
785-650 (95.4%)
Wk-25781
(Cat. No.
JFB.088.1)
Budibudi
Atoll
Conus sp.
(C)
2.98±0.2 1251±30 785-700 (68.2%)
850-670 (95.4%)
Wk-31234
(Cat. No.
E15596-B)
Wanigela
(Rainu)
Conus sp.
(C)
3.0±0.2 944±31 540-485 (68.2%)
600-455 (95.4%)
AA-25130
( Bickler
5)**
Woodlark
(Budibudi
Atoll)
Conus sp.
(C)
4.2 1205±45 750-662 (68.2%)
882-632 (95.4%)
Marine
modelled age
[Rg(t)]
!R (14C yr)
[Rs(t) - Rg(t)]
Wk-26249
[NMNZ
212226]*
Samarai,
Nov 1938
Barbatia sp.
(FF)
2.70±0.2 485±34 459±23 26±34
Wk-21066
[AM:C1191
61]*
Kiriwina
Island
Oct 1944
Pinctada
imbricata
(FF)
-
1.19±0.2
507±17 463±23 44±17
# FF = filter-feeding bivalve; C = carnivore. Hogg et al. (1998) recommend filter feeders for 14C analysis.
* Museum acquisition number: NMNZ = National Museum of New Zealand; AM = Australian Museum;
JFB = Jolika collection, de Young Museum, San Francisco.
** Armshell still in kula, date is from offcut.
Figure 30: The calibrated dates for the four engraved shell artifacts
30
Placing the engraved Conus artifacts in prehistory
by Glenn Summerhayes, Simon Bickler, Pamela Swadling and Harry Beran
To understand when the engraved Conus artifacts were produced it is necessary to give a broad-brush view of
the deep past of the northern Massim and its links with the New Guinea mainland. Three phases stand out in the
northern Massim. In the first two phases there are strong links between the groups living along the northern part
of the eastern tip of New Guinea and the islands of the northern Massim. In the third and final phase strong
trade contact with the New Guinea mainland ceases and is replaced by inter-island trade.
Within the northern Massim the defining features of the first two phases are changes in burial practices, whereas
the last is defined by changes in social networks. Each phase gives some insights into the socio-economic
changes that have occurred in this region. The Conus artifacts are associated with the middle of these phases.
In Phase 1 burials are placed within megalithic stone structures located within villages. This practice maintained
close relationships with the dead. Building these structures required considerable effort and social networking
as even in the Trobriand Islands they included stone slabs quarried on Woodlark Island (Austen 1939/40; Egloff
1979:107; Bickler 2006). The oldest pottery for this phase has been found on Illamu Island located between
Goodenough and Fergusson Islands in a site dating from 1520-1320 calBP / AD 430-630 (Vincent Kewibu pers.
comm. 2007). This pottery includes pedestal vessels and has distant affinities with Lapita forms.
In Phase 2 social attitudes changed, instead of burying the dead within special structures within the village,
bodies were probably buried within villages and their bones were generally subsequently collected and placed
away from villages in caves, islets or other locations (Egloff 1979:107; Bickler 2006; Vere and Young 1980).
Within the northern Massim this phase has come to be called the Expansion Phase and dates from c 1000-500
BP/ c AD 950-1450 (Egloff 1972:155; 1979; commencement date revised Brian Egloff pers. comm. 2011). It is
a period of growth marked by an increase in the number and visibility of archaeological sites. The Expansion
Phase is when engraved Conus artifacts were produced. The three finds from Budibudi Atoll are mid-phase as
they date from 880-630 calBP / AD 1070-1320, whereas the find from Rainu at 600-455 calBP / AD 1350-1495
dates towards its end.
31
Figure 31: Map showing the extent of the social networks operating during the Expansion Phase as well as the
extent of the kula circuit in the 1910s The origin of the engraved Conus artifact circulating in the kula collected on Fergusson Island is not known. Expansion Phase pottery
has been found at Eroro and Boiani, but the extent to which this pottery was traded inland on the New Guinea mainland is not known.
The widespread nature of the pottery we now know belonged to the Expansion Phase was first noted by Leo
Austen (1939/40) when he noticed similarities between Collingwood Bay prehistoric pottery and potsherd
scatters in the Trobriands. This connection was empirically confirmed by Con Key’s (1968) petrographic work.
Key found that the pottery from mainland Collingwood Bay is made from clay collected from the lower
southern slopes of Mt Victory, 61 metres above sea level and about four kilometres from Wanigela area villages
on the shore (Key 1968:653). Brian Egloff (1978) undertook a rudimentary chemical analysis on some
Expansion Phase pottery from the Trobriands, and argued that it was manufactured from clays obtained from
both New Guinea (Collingwood Bay) and Goodenough Island. This pattern of pottery movement is different
from the ethnographic pattern of pottery entering this part of the kula from the Amphlett Islands. On the basis of
these results, Egloff (1978) argued for a wider system of exchange than in the ethnographic kula with the basic
elements and geographical boundaries of modern kula (including the onset of pottery production in the
Amphlett’s) occurring sometime after AD 1500.
Figure 31 shows that Expansion Phase pottery is distributed from Dyke Ackland Bay, Collingwood Bay, and
Goodenough Bay on the mainland, as well as on Goodenough, Nuamata and the Trobriand Islands. The specific
similarities are those that Austen first noted in the shape and decoration of this pottery. The pots he saw were
globular in shape with everted rims, and had grooved decoration, both linear and curvilinear, the latter including
scroll motifs (Lauer 1971:206 after Austen 1939). Bowls are also found and some of these have appliqué spirals
(Austen 1939: Figure 7 and 8). During this phase on Woodlark Island pottery was not imported but made
locally (Bickler 1998: 181).
32
Egloff did not find engraved Conus artifacts, but he did find Collingwood Bay pottery with curvilinear
decorations in his excavations at Wanigela (Brian Egloff pers. comm. 2011). However, both were found in the
mounds that were levelled to infill swampy ground at the Wanigela mission station in the first years of the 20th
century. Pöch (1907a), Seligman and Joyce (1907) and Monckton (1905; 1922) all illustrate this type of pottery
as well as engraved Conus shells. Egloff attributes an age of at least 700 years to this type of pottery based on
the Wanigela sequence. Its production at Wanigela had ceased by the commencement of the Refuge Phase in c
500 BP (Egloff 1979:103).
By c 500 BP the social networks extending out of the northern Massim were no longer primarily orientated to
the New Guinea mainland (Egloff 1978). Instead in Phase 3 the islands of the northern Massim became part of
an increasingly specialised insular trading system. New sources of supply and social networks emerged within
this island world. The Amphlett Islanders now monopolised the supply of pottery to the potting clay deficient
Trobriand Islands (Egloff 1978; 1979), whereas prior to 500 BP this utility had in part been supplied from
Collingwood Bay on the mainland. Likewise Tubetube, Wari and Paneati emerged as trader middlemen in the
last 400-500 years (Irwin 1983:70; Negishi and Ono 2009; 36,44,48).
A number of artifacts, including Conus armshells, are traditionally used in community exchanges associated
with marriage, death and rights to land in the Massim. Within the Trobriands Conus valuables used in this way
are classed as kitoum/ kitomu, but as required can be placed in the kula (Berde 1983:441; Damon 2002:119;
Weiner 1983:161). Those armshells (mwali) in the kula are ceremonially exchanged for necklaces (soulava).
Malinowski (1922) documents how armshells circulate within the kula in an anti-clockwise direction, whereas
necklaces travel in a clockwise direction. The exchange of one item for the other is not a simultaneous
transaction, there is a delay, as each has to be sought at the residence of the reciprocator. Figure 31 shows the
extent of the kula circuit as recorded by Malinowski in the 1910s. Outside this circuit, in other parts of the
Massim as well as up the Papuan coast, armshells and necklaces are also traded. In these areas there is no
requirement that one be exchanged for the other (Leach 1983).
Whether the engraved Conus artifacts were exchanged for another specific item, such as necklaces as in the
kula, is not known.
Concluding remarks
by Pamela Swadling and Harry Beran
Engraved Conus artifacts are signature artifacts of the Expansion Phase, c 1000-500 BP/ c AD 950–1450
(Egloff 1972:155; 1979; Brian Egloff pers. comm. 2011). The 32 finds reported here show that social networks
extended from Collingwood Bay on the New Guinea mainland across the northern Massim to Budibudi Atoll, a
distance of some 500 km. Previously these prehistoric links had been demonstrated to extend only as far east as
the Trobriand Islands with the presence of pottery from Collingwood Bay and Goodenough Island in the
Trobriand Islands. The engraved Conus artifacts indicate that these networks extended east as far as Budibudi
Atoll. Mainland and Goodenough pottery was not needed there as pottery was produced on Woodlark for local
use. The network link with Collingwood Bay lasted until c 500 BP when Massim social networks became more
island-oriented.
It is remarkable that 27 of the 32 shells located so far have been found in just two localities; namely, Wanigela
and Budibudi. This may be partly due to their preservation in the middens at Rainu in the Wanigela area and
under coral-rock mounds in Budibudi. And partly it may be due to the tendency of heirloom objects that go out
of ‘fashion’ to be traded to a peripheral area where they are still valued. It is not known when they were cached
under the coral-rock mounds in Budibudi. Very little archaeological work has been done on the New Guinea
mainland or in the Massim region. Hence, it is possible that significant numbers of engraved Conus shell
33
artifacts may be found in future in southeast Papua New Guinea.
This study also provides a chronological framework for another representation of curvilinear art in this region,
namely the rock carvings of Goodenough Bay (Egloff 1970). In this respect it is probably significant that
Expansion Phase potsherds have been found at Boiani in Goodenough Bay (Egloff 1979:111; Brian Egloff pers.
comm. 2011).
The Conus shell artifacts discussed in this essay provide the earliest radiocarbon dated representation of the
Massim art style, one of the major art styles of Papua New Guinea. This style was first documented by A.C.
Haddon in his book The Decorative Art of British New Guinea (1894) and continues to be produced by Massim
woodcarvers. The engravings made on the Conus shells indicate that this curvilinear style has been carved for at
least 800 years.
Acknowledgements
We owe special thanks to John Kasaipwalova, who told us about three engraved shells in the Massim region,
Jim Specht, who advised us of the engraved shell in the Horniman Museum and Brian Egloff for his generous
comments and advice on dating the Expansion Phase.
We are indebted to the following for assisting with the inspection of engraved shells in museums or providing
important information: Richard Aldridge, Chris Boylan, Yvonne Carillo, Colin Davidson, Jack Fenner, Rebecca
Fisher, John Friede, Rachel Hand, John Heaton, Christina Hellmich, Jutta Malnic, Anthony J.P. Meyer, Wayne
Modest, Sue O’Connor, Robin Torrence, Peter White and Gabrielle Weiss.
Bibliography
Austen, L. 1939-40. Megalithic Structures in the Trobriand Islands. Oceania 10: 30-53.
Barker, J. 2008. Ancestral lines: the Maisin of Papua New Guinea and the fate of the rainforest. Peterborough:
Broadview Press.
Bellwood, P. 1979. Man’s Conquest of the Pacific: The Prehistory of Southeast Asia and Oceania. New York:
Oxford University Press.
Beran, H. 1988. Betel-chewing Equipment of East New Guinea. Aylesbury: Shire.
Beran, H. and E. Aguirre 2009. The Art of Oro Province: A Preliminary Typology. Sydney: Oceanic Art
Society. [Published on C.D.]
Bickler, S.H. 2006. Prehistoric stone monuments in the northern region of the Kula Ring. Antiquity 80: 38-51.
Bickler, S.H. 1998. Eating stone and dying: Archaeological survey on Woodlark Island, Milne Bay Province,
Papua New Guinea. Unpublished PhD. University of Virginia.
Bronk-Ramsey, C. 2010. OxCal Program v4.10, Radiocarbon Accelerator Unit, University of Oxford.
Calcinai, B., A. Arillo, C. Cerrano and G. Bavestrello 2003. Taxonomy-related differences in the excavating
micro-patterns of boring sponges. Journal of the Marine Biological Association of the United Kingdom 83: 37-
39.
34
Campbell, S. F. 2002. The Art of Kula. Oxford: Berg.
Chappell J. and H.A. Polach 1976. Holocene sea-level change and coral-reef growth at Huon Peninsula, Papua
New Guinea. Geological Society of America Bulletin 87: 235-240.
Chignell, A.K. 1911. An outpost in Papua. London: Smith, Elder & Co.
Damon, F. H. 2002. Kula Valuables: the Problem of Value and the problem of Names. L’Homme 162: 107-36.
Edwards, R.L., J.W. Beck, G.S. Burr, D.J. Donahue, J.M.A. Chappell, A.L. Bloom, E.R.M. Druffel, F.W.
Taylor. 1993. A large drop in atmospheric 14C/12C and reduced melting in the younger Dryas, documented
with 230Th ages of Corals. Science 260: 962-968.
Egloff, B. 1970. The rock carvings and stone groups of Goodenough Bay, Papua. Archaeology and Physical
Anthropology in Oceania 5: 147-156.
Egloff, B. 1971. Collingwood Bay and the Trobriand Islands in recent prehistory. Canberra: PhD thesis
Australian National University.
Egloff, B.J. 1972. The sepulchral pottery of Nuamata Island, Papua. Archaeology and Physical Anthropology in
Oceania 7: 146-163.
Egloff, B.J. 1978. The Kula before Malinowski: a changing configuration. Mankind 11(3): 429-435.
Egloff, B.J. 1979. Recent Prehistory in southeast Papua. Terra Australis 4. Department of Prehistory, Research
School of Pacific Studies. Australian National University, Canberra.
Gash, J.L. 1975. Prehistoric pottery remains at Wanigela, 1905. In Whittaker, J.L. (ed), Documents and
readings in New Guinea history: pre-history to 1889. Milton: Jacaranda: 159-161.
Grunne, B. de. 1979. Art Papou. Brussels: louis musin éditeur.
Haddon, A.C. 1894. The decorative art of British New Guinea: a study in Papuan Ethnography. Dublin, Royal
Irish Academy, Cunningham memoirs 10.
Hamson, M. and R. Aldridge 2009. Art of the Massim & Collingwood Bay. Palos Verdes Estates, Cal.: Michael
Hamson OCEANIC ART.
Hogg, A, T.F.G. Higham, J. Dahm 1998. Radiocarbon dating of modern marine and estuarine shellfish.
Radiocarbon 40: 975-984.
Irwin, G.F. 1983. Chieftainship, kula and trade in Massim prehistory. In Leach J.W. and E. Leach (eds), The
Kula: New perspectives on Massim exchange. Cambridge: Cambridge University Press: 29-72.
Key, C.A. 1968. Pottery manufacture in the Wanigela area, Collingwood Bay, Papua. Mankind 6: 653-657.
Lauer, P. 1971. Changing Patterns of Pottery Trade to the Trobriand Islands. World Archaeology 3: 197- 209.
Leach, J. 1981. Imdeduya: a Kula folktale from Kiriwina. Bikmaus: Journal of Papua New Guinea Affairs, Ideas
and the Arts.
35
Leach, J. 1983. Introduction. In Leach J.W. and E. Leach (eds), The Kula: New perspectives on Massim
exchange. Cambridge: Cambridge University Press: 1-26.
Liep, J. and D. Affleck 1983. ‘This Civilising Influence': The Colonial Transformation of Rossel Island Society.
The History and Anthropology of the Massim, Papua New Guinea, Part 2. The Journal of Pacific History 18
(2): 113-133.
Mackay, R. D. 1971. An Historic Engraved Shell from the Trobriand Islands, Milne Bay District. Papua and
New Guinea Public Museum and Art Gallery Records 1 (1): 47-51.
Malnic, J. with J. Kasaipwalova 1998. Kula: Myth and magic in the Trobriand Islands. Wahroonga: Cowrie
Books.
McGregor, H.V., M.K. Gagan, M.T. McCulloch, E. Hodge, G. Mortimer 2008. Mid-Holocene variability in the
marine 14C reservoir age for northern coastal Papua New Guinea. Quaternary Geochronology 3: 213-225.
Meyer, A.J.P. 1995. Oceanic Art. Köln: Könemann.
Monckton, C.A.W. 1905. Resident Magistrate’s report: North-eastern Division, British New Guinea. In British
New Guinea Annual Report for the year ending 30th June 1904, Government Printer, Melbourne, Appendix D:
31-34.
Monckton, C.A.W. 1922. Last Days in New Guinea: Being Further Experiences of a New Guinea Resident
Magistrate. London: John Lane The Bodley Head.
Negishi, Y. and R. Ono 2009. Kasasinabwana shell midden: The prehistoric ceramic sequence of Wari Island in
the Massim, Eastern Papua New Guinea. People and Culture in Oceania 25: 23-52.
Newton, D. 1975. Massim: Art of the Massim area, New Guinea. New York: The Museum of Primitive Art.
Newton, D. 1988. Reflections in Bronze: Lapita and Dong-son art in the Western Pacific. In J. P. Barbier and D.
Newton (eds) Islands and Ancestors: Indigenous Art Styles of Southeast Asia. Munich: Prestel.
Petchey F., M. Phelan, P. White 2004. New "R values for the southwest Pacific Ocean. Radiocarbon, 46: 1005-
1014.
Petchey F., A. Anderson, A. Zondervan, S. Ulm, A. Hogg 2008. New Marine "R values for the South Pacific
Subtropical Gyre Region. Radiocarbon 50(3): 373-397.
Pöch, R. 1907a. Einige bemerkenswerte Ethnologika aus Neu-Guinea. Mitteilungen der Anthropologischen
Gesellschaft in Wein 37: 57-71.
Pöch, R. 1907b. Travels in New Guinea, British and Dutch New Guinea. Geographical Journal 30 (6): 609-
616.
Pöch, R. 1907c. Ausgrabungen alter Topfscherben in Wanigela (Collingwood Bay). Mitteilungen der
Anthropologischen Gesellschaft in Wien 37: 137-139.
36
Reimer, P.J., M.G.L. Baillie, E. Bard, A. Bayliss, J.W. Beck, P.G. Blackwell, C. Bronk Ramsey, C.E. Buck,
G.S. Burr, R.L. Edwards, M. Friedrich, P.M. Grootes, T.P. Guilderson, I. Hajdas, T.J. Heaton, A.G. Hogg, K.A.
Hughen, K.F. Kaiser, B. Kromer, F.G. McCormac, S.W. Manning, R.W. Reimer, D.A. Richards, J.R. Southon,
S. Talamo, C.S.M. Turney, J. van der Plicht, C.E. Weyhenmeyer 2009. Radiocarbon 51: 1111-1150.
Rützler, K. 1975. The Role of Burrowing Sponges in Bioerosion. Oecologia 19: 203-216.
Schönberg, C.H.L. 2002. Substrate Effects on the Bioeroding Demosponge Cliona orientalis. 1. Bioerosion
Rates. Marine Ecology 23(4): 313-326.
Seligman, C.G. 1909. A Type of Canoe Ornament with Magicial Significance from South-eastern British New
Guinea. Man 9: 33-5 and Plate C.
Seligman, C.G. and T.A. Joyce 1907. On prehistoric objects in British New Guinea. In Rivers, W.H.R. et al
(eds) Anthropological Essays presented to Edward Burnett Tylor, in honour of his 75th birthday. Oxford:
Clarendon: 325-41.
Senft, G. 2008. The Teachings of Tokunupei. In J. Kommers and E. Venbrux (eds) Cultural Styles of
Knowledge Transmission. Amsterdam: aksant: 139-44.
Specht, J. 1988. Pieces of Paradise. Australian Natural History, Supplement No. 1.
Steinberg, C.R., S.M. Choukroun, M.M. Slivkoff, M.V. Mahoney, R.M. Brinkman 2006. Currents in the
Bismarck Sea and Kimbe Bay, Papua New Guinea. Australian Institute of Marine Science and The Nature
Conservancy. TNC Pacific Island Countries Report N. 6/06.
Stuiver, M., G.W. Pearson, T.F. Braziunas 1986. Radiocarbon age calibration of marine samples back to 9000
cal yr BP. Radiocarbon 28: 980-1021.
Tets, G.F. van 1971. New view on mystery engraving. Australian Natural History 17 (2): 58.
Vere, M. de and M.W. Young 1980. Secondary burial on Goodenough Island, Papua: Some archaeological and
ethnographic observations. Occasional Papers in Anthropology 10: 229-248. St Lucia: University of
Queensland.
Vermeij, G.J. 1978. Biogeography and Adaptation: Patterns of Marine Life. Cambridge MA and London:
Harvard University Press.
Walker, S.E. 1998. Endobionts on Modern and Fossil Turritella from the Northern Gulf of California Region.
Ichnos 6: 99-115.
Weiner, A.B. 1983. ‘A world of made is not a world of born’: doing kula in Kiriwina. In Leach J.W. and E.
Leach (eds), The Kula: New perspectives on Massim exchange. Cambridge: Cambridge University Press: 147-
170.
White, J.P., H.J. de S. Disney and J.C. Yaldwyn 1970. Prehistoric Papuan Engraving. Australian Natural
History June: 344-345.
37
Recommended