'Stoning fish?' A hitherto unrecorded class of stone artefact from the coastal Pilbara.

Akerman, Kim

Abstract: Yodda-like stone artefacts from the coastal Pilbara region of Western Australia differ markedly from recorded yoddas and constitute a hitherto unrecorded Aboriginal Australian stone implement. I suggest that the implements were possibly used as missiles for killing or stunning fish.

**********

For many years I have been engaged in examining the distribution of yoddas--a rare and unusual Australian tanged stone implement. In June 2002 while examining yoddas in the Western Australian Museum (WAM), I was shown a collection of 19 tanged implements that resembled crude, short-handled table-tennis bats, which had been donated to the WAM in 1995. These artefacts formed part of a larger collection of stone and shell implements, originating from a site located on the mouth of Cowrie Creek, which lies between Cape Cossigny and Cape Thouin on the Pilbara coast.

The collection had been made by Gwen and Snowy Collison of Gilgandra, New South Wales, in July 1973. According to the information provided, the artefacts had been found in an area of swale, said to be covered with hearths, midden material and artefacts. I was able to contact Mrs Collison, who subsequently provided more data about the collection and a photograph showing a further five tanged objects. These latter artefacts, part of the 'Australian Collection', are now in the Gilgandra (NSW) Shire Museum. The Collisons had collected the tanged artefacts, believing them to be yoddas.

In late 2002, I examined photographs of the five yoddas held in the South Australian Museum (SAM) collections. One (A62506), was said to be from Gilgandra, NSW. It had been donated by I Collison of Gilgandra and was said to have come 'from the grave of Miranda, chief of the Aberombie [sic] tribe' (presumably this should be Abercrombie = Ngadjuri). It had come to the SAM prior to 1971. I recognised it as one of the Cowrie Creek tanged implements and contacted Mrs Collison again to query the information provided by the SAM. Mrs Collison denied that they had ever collected 'yoddas' in New South Wales and thought that her husband had forwarded an example of the Pilbara artefact to the SAM. It appears that a mix-up had occurred in terms of location and time of donation, when registering it at the SAM.

The Cowrie Creek artefacts

Examination quickly revealed that the Cowrie Creek artefacts, although superficially yodda-like, appeared to represent a separate class of implement in their own right. Rather than having a relatively narrow tang in relation to the overall width of the artefact, as with yoddas, the tangs on these implements were generally rectangular to square in shape. Again, unlike tangs on yoddas, which generally make up at least half the total length of the artefact, tangs of the Collison artefacts were usually less than a third of their total length. A further factor was that, rather than being shaped like a bicycle seat as yoddas are, the artefacts in the Collison Collection resemble short-handled table-tennis bats. Most have a circular proximal end, although some are sub-rectangular and a few have obvious marginal damage that affects their overall symmetry (Figure 1).

[FIGURE 1 OMITTED]

Unlike yoddas, which are generally rather massive artefacts, the Collison implements are relatively thin in relation to their length/width proportions, reinforcing their bat-like appearance. Thirteen examples are very symmetrical about their longer axis. Of those that are not quite as symmetrical, it appears that the manufacturer has merely selected a suitably sized and relatively flat piece of reef float and simply created a tang at one end, rather than carefully shaping the entire piece. The symmetry of B2955-3 is affected by a diagonal break that appears to have removed a triangular piece of stone that has reduced the area of the tang by about half. Tables 1 and 2 provide metrical and other data for the Collison Collection and known Australian yoddas, respectively.

Both igneous and sedimentary rocks have been used to provide the raw material for artefacts in the Collison Collection. The sedimentary rocks appear to have been derived from the adjacent marine environment, whereas the hard stone would have had to have been transported at least tens of kilometres to the site. The majority, however, are made from calcareous shingle picked up from an adjacent reef. Many have been bored through by marine bivalves, such as those belonging to the families Petricolidae and Mytilidae (sub-family Lithophaginae) for example, while others have traces of calcareous worm tubes over their surfaces. One example, B2995-8, still has the paired valves of the crevice-nestling bivalve Irus irus in situ within holes in the stone.

Of the hard stone artefacts, five appear to have been made from fragments of granite grindstones. The specific nature of these pieces--texture, crystal size and alignment, flow structures, and so on--differs from one to the other, suggesting that they were not made from the same original grindstone. One of the granite pieces appears to be derived from a natural exfoliation spall. The other hard stone piece (B2955-15) appears to be a piece of fine-grained hornfels. There has been virtually no attempt to bring this to a symmetrical shape, although one end has been reduced to form a distinctive grip.

Of the Gilgandra collection of five tanged tools, three appear to be made of granite, one of shelly sandstone and one of dolerite, possibly an exfoliated cortical spall. The South Australian Museum example is made from a piece of granite.

Flaking patterns created during manufacture are very clear on the hard stone tools, whereas weathering has all but obliterated traces of manufacture on those made of sedimentary rock. Generally, the examples made from hard stone are the neatest in terms of overall symmetry. Regardless, it appears that either shingle or spalls of hard stone have been reduced to the desired form by marginal flaking and crushing, possibly while resting on an anvil. The granite example shows signs of weak abrading about the 'handle', making it more comfortable to hold and use.

My own experiments in the manufacture of these implements, using exfoliated sheets of granite as raw material, were more successful when the material was rested on an anvil rather than flaked freehand. Using a suitably thin exfoliated sheet of granite it took no more than 90 seconds to create a small tang and reduce the margins to an appropriate shape and size.

In order to appreciate the character of the Cowrie Creek artefacts it is necessary to provide some details on the artefact that both professional and amateur archaeologists have confused them with--the yodda.

Discussion of yoddas

In 1934, Casey, describing a series of unusual tanged implements found in New Guinea and Australia, unwittingly lumped them together on the basis of similar plan view. McCarthy further confused the issue when he described yoddas generally as a 'tanged axe' (1940:43) and concluded that they were related to the shouldered or butt-reduced edge-ground axes of western-central Queensland. Maintaining the assumption that the Australian examples were closely related to the Papuan specimens, McCarthy (1940:45-6) also speculated that they reflect an externally derived Bronze Age influence. Later, he suggested that yoddas represent a culture element introduced into Australia from New Guinea via the Torres Strait (McCarthy 1941:371).

The name 'yodda' was given to these artefacts by McCarthy (1944:265) after the location in Papua New Guinea where the original tanged specimen had originated. The New Guinea 'type specimen' is a tanged implement with convex cutting edge set normal to the tang and which is made on a large flake of obsidian (Seligman 1915; Seligman & Joyce 1907:327-8). Apart from a broad similarity in terms of the outline plan view, this piece bears little resemblance to the Australian implements. In its proportions the PNG example can be seen to have a relatively short, thin tang in relation to its overall length and is one of few examples of a 'yodda' that is wider than it is long (Table 2). Using Groube's (1986:168-9) classification, this artefact, the 'Type' specimen, is in fact a stemmed axe.

Australian yoddas are large tanged implements with curved distal end and lenticular section that range in size from 170 to 300 mm in length and between 1000 and 5000 g in weight. Some authors (Casey 1934:94, 1936:92; McCarthy 1941:371; McCourt 1975:108) have described them as resembling stone bicycle (push-bike) saddles or seats. Most appear to have been made by carefully flaking tabular, hard stone material, some are refined by hammer-dressing and a few are said to be ground on the edges or even across the body of the implement. Not withstanding that very few ground examples have been discovered, McCarthy and his colleagues (1946:77) suggested that the flaked but unground examples are unfinished blanks. Later, McCarthy (1976:55) suggested that they are monolithic axes or hatchets, with head and helve made from a single piece of stone.

The use of the term 'tang' should perhaps be restricted to a hafting mechanism, and Australian monolithic 'tanged' implements are more correctly provided with a handle element for grasping directly. I have, however, elected to retain the term 'tang' in this article, with the understanding that I refer to a handle. Golson (2001), discussing tanged and waisted artefacts from New Guinea and Australia, failed to mention the Australian yoddas. Consequently, implements where the 'tang' is the handle or helve are ignored in his discussion of waisted and butt-modified artefacts. In fact, yoddas could be described as double-bitted monolithic stone axes or hatchets.

McCarthy limited the distribution of yoddas to New South Wales and Queensland, with the exception of a single possible example, described as a yodda blank, from Milingimbi in Arnhem Land; Casey (1934), however, records their presence in South Australia and Gippsland. It is now known that they also occur in Western Australia, examples having been found in three regions: the south-west, the Pilbara, and the Kimberley.

Yoddas are not a common artefact and McCarthy stated that only about 20 examples had ever been collected across the continent. It is now known that they are more common. I am aware of at least four previously unrecorded examples and that at least three examples have been found in the Kimberley and two in the Pilbara. They are, however, still a rare form of artefact and to date they have only been collected as isolated finds. None has been found in excavations or dated deposits and consequently cannot be ascribed to any particular time in Australia's past. A large collection of yoddas from a single site would require a reconsideration of the status of the implements and perhaps shed some light on their function.

Speculations on the function of the Cowrie Creek tanged implements

Examination of the ethnographic records does not overly assist in ascribing a possible function to the Cowrie Creek tanged implements. However, their relatively high numbers on a single site (in contrast to the single finds of yoddas) and the variety of lithologies used to make them, including the use of expediently gathered and minimally shaped marine shingle, suggest that they fulfilled a rather mundane niche in the lives of the people who made and used them.

My own view is that they were missiles, and most probably associated with fishing strategies akin to the use of specialised boomerangs for fishing.

Fishing with boomerangs has been reported for the north-western coast of Western Australia (Clement 1903:3-4), from Roebourne and Cossack to Cape Leveque, the northernmost extremity of the Dampierland Peninsula. Peggs (1903:326, 366, plate XV) reported on and illustrated fishing boomerangs from Broome. Smith and Kalotas (1985:343, fig. 8) identified woods used by the Bardi of the Cape Leveque area for making fishing boomerangs and also (plate) showed one being used by a boy. Until 1970, metal boomerangs, called tangks, because they were often cut from discarded galvanised water-tanks, were commonly used by Aboriginal men and boys to kill fish along the Eighty-Mile Beach-Cape Leveque coast.

My own experiments with the stone bats involved throwing a replicated example into the water, while aiming at objects between 1.5 and 3 m away, and showed that they were efficient missiles. The stone was gripped by the small handle which is big enough to provide purchase and thrown edge-on at the intended target. The mass of the stone, coupled with a relatively small striking area (the edge of the artefact) meant that they could be thrown at high velocity while meeting less drag on contact with the water, and strike with greater impact than occurs when using a wooden fishing-boomerang. The handle also assisted in controlling accuracy. They are not as efficient, however, as the post-contact boomerangs cut from heavy-gauge, galvanised iron sheets. The thinness of the metal, coupled with the mass of the material and the length of the arms (which by covering a greater area to contact provided more latitude in aiming the weapon), made metal boomerangs efficient shallow-water fishing tools. Fish weighing up to 3 kg could be killed or stunned, and were regularly taken using these tools.

Basedow (1929:130-1) stated, in a frustratingly brief sentence: 'A constant watch is kept for large fish, which may be swimming close to shore, in order that they might be dispatched with a stone or a throwing stick'. No specific location is provided for this practice. The Piddingtons (1932:345) reported on the panaing method of fishing practised by the Karajarri of La Grange Bay between Port Hedland and Broome. This method of fishing is practised at low tide, the fishers wading out and killing fish with spears or with tyimbi, flat wooden implements 'resembling in shape a large squash racket'. Interestingly the Piddingtons did not refer to the use of the stone fish-traps (gurijungu) found in the area, and neither they, nor Petri-Odermann (1963), referred to boomerang fishing in the La Grange area. According to my own information, boomerang fishing was certainly practised there, and Pat Vinnicombe had photographed a piece of sheet iron on a La Grange site from which a fishing boomerang had been cut (Vinnicombe, pers. comm., 2002).

There is a further possibility that the stone bats were used not only to directly kill fish but also to assist in directing them into nets and traps. I have recorded the use of natural stones to direct fish into pangala, stone fish-pounds, by the Bardi and Nyul Nyul (Akerman 1976). Net-fishing was extensively undertaken along the Pilbara coast (Clement 1903:3-4; Gregory 1884:58, 72-3; Withnell 1965:21-2), although there is no record of driving fish. Throughout the twentieth century, flat pieces of wood, 'tin' and slate, as well as tin boomerangs, were used to kill fish by many boys who lived along the Swan River. As one anonymous contributor noted, 'The kierlie-throwers have gone, but their art lives on in the white youngsters of today, who use flat pieces of wood or, preferably tin, or, best of all (if they can get it) a thin sheet of slate in the form of a rough triangle for this exacting method of fishing' (Anon. 1950).

Interestingly, Burdett-Scougall (1949) reported that Basques used bat-shaped missiles, not for fishing but in the same manner as some round and triangular pieces of bark are reported to have been used in conjunction with net-hunting of waterfowl by Aborigines (Angas 1969, vol. 1:100; Eyre 1964, vol. 2:287-8). Called zimbela, the bats are hurled above flocks of pigeons traversing valleys in the Pyrenees, causing them to dive low and, with luck, fly into concealed nets strung across their flight path. One caption to a photograph (Burdett-Scougall 1949:406) reads: With split second timing the zimbela tosser throws his paddle-like bats above an approaching flock. Twirling like boomerangs and making a swishing sound, the disks imitate hawks in flight. This causes pigeons to fly low for fear of falcon attacks from above. Hunters time their actions so that pigeons are flying but a few feet above ground when they reach the netted ambushes.

I present this as a further instance where bat-shapes have been considered a suitable form for a missile. I had also considered the possibility that the stone bats may fall into that class of weapon found in Polynesia and best exemplified by the Maori patu, or stabbing club. However, the low masses and fragile nature of some of the implements, coupled with the extremely short handles, preclude such a function.

Conclusion

I have attempted to distinguish a new form of stone tool that, while superficially similar to a recognised form, the yodda, warrants consideration in its own right. I suggest that these artefacts are related to an everyday economic activity--fishing. I recognise that my ideas are based on a single collection from one site and that much more needs to be done in determining the distribution, particularly within the Pilbara region, of the 'fishing stones' before my interpretation can be ratified. It is my aim to draw the attention of researchers to this artefact and also to provide more recent data on the distribution of yoddas in Australia. Table 1 Lithology and dimensions of the artefacts in the Collison Collection, Western Australian Museum (B2955:1-19); in the Australian Collection, Gilgandra Council, NSW (G1-6); and in the South Australian Museum (A62056) Museum Length Width reg no. Stone type (mm) (a) (mm)(b) 1 B2955-1 granite (c) 200 160 2 B2955-2 granite (c) 180 145 3 B2955-3 granite (c) 165 110 4 B2955-4 granite 230 145 5 B2955-5 granite (c) 165 125 6 B2955-6 granite (c) 160 120 7 B2955-7 granite gneiss (c) 205 134 8 B2955-8 ferruginous shelly 199 150 marl (d) 9 B2955-9 shelly sandstone 165 150 10 B2955-10 shelly sandstone (d) 212 185 11 B2955-11 sandy calcretele (e) 220 165 12 B2955-12 calcrete (e) 180 130 13 B2955-13 gritty sandstone (d) 170 152 14 B2955-14 friable coarse quartz 195 140 conglomerate (d) 15 B2955-15 fine-grained 116 113 meta-sedimentary rock 16 B2955-16 cross-bedded calcareous 160 110 sandstone (d) 17 B2955-17 ferruginous shelly 185 145 marl (d) 18 B2955-18 shelly sandstone 235 225 19 B2955-19 sandy limestone (e) 225 156 20 G1 granite 175 150 21 G2 granite 240 210 22 G3 granite 205 155 23 G4 granite 160 150 24 G5 granite 145 120 25 G6 tabular dolerite 260 210 26 A62056 granite 215 170 Museum Thickness Tang L/W Mass reg no. Stone type (mm) (mm) (g) 1 B2955-1 granite (c) 20 50/70 753 2 B2955-2 granite (c) 20 40/55 714 3 B2955-3 granite (c) 24 40/70 681 4 B2955-4 granite 12 70/90 732 5 B2955-5 granite (c) 18 * 40/50 518 6 B2955-6 granite (c) 18 40/60 462 7 B2955-7 granite gneiss (c) 30 70/90 1197 8 B2955-8 ferruginous shelly 15 50/80 643 marl (d) 9 B2955-9 shelly sandstone 28 40/70 859 10 B2955-10 shelly sandstone (d) 38 55/75 1161 11 B2955-11 sandy calcretele (e) 25 75/80 949 12 B2955-12 calcrete (e) 10 40/70 254 13 B2955-13 gritty sandstone (d) 25 40/70 586 14 B2955-14 friable coarse quartz 45 66/70 903 conglomerate (d) 15 B2955-15 fine-grained 22 50/80 619 meta-sedimentary rock 16 B2955-16 cross-bedded calcareous 30 50/60 485 sandstone (d) 17 B2955-17 ferruginous shelly 25 40/60 645 marl (d) 18 B2955-18 shelly sandstone 35 65/80 1633 19 B2955-19 sandy limestone (e) 35 45/70 961 20 G1 granite 20 35/60 700 21 G2 granite 25 60/75 1500 22 G3 granite 30 55/75 1100 23 G4 granite 20 34/75 780 24 G5 granite 20 40/60 500 25 G6 tabular dolerite 30 50/70 2000+ 26 A62056 granite 35 -- -- Notes: (a) length along axis of the tang; (b) width taken normal to the axis of the tang; (c) fragment of grindstone; (d) stone bears worm tubes or evidence of activity of rock-boring bivalves; (e) artefact damaged or broken in some way. Table 2 Recorded dimensions of the Papuan Type Specimen and known Australian yoddas. (Dimensions converted to metric, or estimated, from the original publications) Surface treatment Specimen Stone type of artefact Yodda Valley obsidian flaked PNG (1) Deighton Vic (2) basal hammer-dressed? Cooyar Qld basalt hammer-dressed? (cast) (2) Kallara, Darling quartzite flaked River NSW (2) (3) Rockhampton Qld (2) basalt ground Brisbane Qld (2) limestone flaked, hammer-dressed? ground Oberon NSW (4) trachyte hammer-dressed Kimberley WA (5, 8) basalt flaked, hammer-dressed? ground ** Mt Buckle Vic (5) m/m s/stone flaked Condamine Qld (5) m/m shale flaked Clayton Creek SA (6) quartzite flaked Idracowra NT (6) quartzite(?) flaked West Kimberley (7) dolerite(?) flaked West Kimberley (7) dolerite(?) flaked Kennedy Ranges, dolerite(?) flaked Pilbara WA (8) Morowa WA (WAM) dolerite Pilbara(?) (WAM) (9) quartzite (?) flaked Sherlock, Pilbara WA diorite flaked (WAM B1987) (9) Norseman, south-west tabular chert flaked WA (WAM B3634) (9) Yodda-like club(?), gneiss flaked Narrogin WA (WAM) (9) Yodda-like club(?), dolerite flaked Miling, south-west WA (WAM 11329) (9) Myponga SA (4) micaceous hammer-dressed (SAM A4552) (10) sandstone Kinchega Station NSW quartzite hammer-dressed (4) (SAM A4551) (10) Coober Pedy SA quartzite (SAM A57424) (10) Lake Windabout SA quartzite (SAM A60102) (10) Mulgaria, Lake Frome quartzite flaked SA (11) Total length Width Specimen (mm) (mm) Yodda Valley 183 (tang = 50) 215 PNG (1) Deighton Vic (2) 331 237 Cooyar Qld 225 244 (cast) (2) Kallara, Darling 317 170 River NSW (2) (3) Rockhampton Qld (2) 197 163 (one edge damaged) Brisbane Qld (2) 270 237 Oberon NSW (4) 262 181 Kimberley WA (5, 8) 270 230 Mt Buckle Vic (5) -- -- Condamine Qld (5) 240 153 Clayton Creek SA (6) 200 125 Idracowra NT (6) 352 175 West Kimberley (7) 250 145 (one edge damaged) West Kimberley (7) 172 141 Kennedy Ranges, 237 160 Pilbara WA (8) Morowa WA (WAM) 240 95 Pilbara(?) (WAM) (9) 295 230 Sherlock, Pilbara WA 275 221 (WAM B1987) (9) Norseman, south-west 206 182 WA (WAM B3634) (9) Yodda-like club(?), 279 140 Narrogin WA (WAM) (9) Yodda-like club(?), 245 142 Miling, south-west WA (WAM 11329) (9) Myponga SA (4) 370 225 (SAM A4552) (10) Kinchega Station NSW 250 180 (4) (SAM A4551) (10) Coober Pedy SA 250 220 (SAM A57424) (10) Lake Windabout SA 270 270 (SAM A60102) (10) Mulgaria, Lake Frome 215 235 SA (11) Thickness Specimen (mm) Mass (gg) Yodda Valley -- -- PNG (1) Deighton Vic (2) -- 5760 Cooyar Qld -- -- (cast) (2) Kallara, Darling -- 3720 River NSW (2) (3) Rockhampton Qld (2) -- 1560 Brisbane Qld (2) 75 -- Oberon NSW (4) 24 1470 Kimberley WA (5, 8) 40 1928 Mt Buckle Vic (5) -- -- Condamine Qld (5) 31 2640 Clayton Creek SA (6) 37 -- Idracowra NT (6) -- -- West Kimberley (7) 50 1539 West Kimberley (7) 50 1041 Kennedy Ranges, 45 1500 Pilbara WA (8) Morowa WA (WAM) 70 2641 Pilbara(?) (WAM) (9) 45 cast Sherlock, Pilbara WA 25 1784 (WAM B1987) (9) Norseman, south-west 34 985 WA (WAM B3634) (9) Yodda-like club(?), 50 2351 Narrogin WA (WAM) (9) Yodda-like club(?), 60 1856 Miling, south-west WA (WAM 11329) (9) Myponga SA (4) 70 6720 (SAM A4552) (10) Kinchega Station NSW 50 2280 (4) (SAM A4551) (10) Coober Pedy SA 270 -- (SAM A57424) (10) Lake Windabout SA 80 -- (SAM A60102) (10) Mulgaria, Lake Frome 60 2747.8 SA (11) * artefact has been subjected to flaking, hammer-dressing or grinding during the manufacturing process. ** artefact described by McCarthy as being smoothed by stream activity or weathering. McCarthy could not determine whether the pitting found on the body of the implement was due to human or natural activity. My own view is that this is a flaked, hammer-dressed and ground piece, the grinding having been effected by a hand-held abrader rather than a static grindstone. Notes: m/m = metamorphosed; s/stone = sandstone; (1) Seligman & Joyce 1907; (2) Casey 1934; (3) Thorpe 1928; (4) Casey 1936; (5) McCarthy 1952; (6) McCourt 1975; (7) Akerman collection; (8) Fr Gil Museum, Kalumburu; (9) WA Museum; (10) SA Museum; (11) collection of Dr Marek Zbik

ACKNOWLEDGMENTS

This article is an amended version of a larger presentation I made to the Skilled Production and Social Reproduction--Aspects on Traditional Stone-Tool Technologies symposium, held at Uppsala, Sweden, in August 2003. I am indebted to Mance Lofgren and Charley Dortch for drawing my attention to the Collison Collection and permitting me to examine and photograph both it and the yoddas held in the collection of the Western Australian Museum. Mrs Gwen Collison generously provided me with further information of the original finds and also sent me images and the dimensions of other Cowrie Creek artefacts now housed in the Gilgandra Shire Museum. Phillip Manning of the South Australian Museum drew my attention to the yoddas held by that institution. The late Dr Pat Vinnicombe generously shared her information on a metal boomerang relic she found on a site near La Grange. Dr Shirley Slack-Smith, also of the Western Australian Museum, kindly identified the bivalves embedded in some of the artefacts and provided details of rock-boring molluscs. Jacqui Ward of the Tasmanian Museum and Art Gallery scanned the images for me. Val Hawkes read and corrected the draft manuscript and tidied up my writing immeasurably. To all these people my deepest thanks. Finally, I am grateful to Graeme Ward and two anonymous reviewers who critiqued the original manuscript.

REFERENCES

Akerman, K 1976, 'Fishing with stone traps on the Dampierland Peninsula, WA', Mankind 10(3):182.

Angas, GF 1969 (1847), Savage life and scenes in Australia and New Zealand, Libraries Board of South Australia, Adelaide (Australiana Facsimile Editions).

Anon. 1950, 'Kierlies for fun', Wild Life 12(3):119.

Basedow, H 1929, The Australian Aboriginal, Preece & Sons, Adelaide.

Burdett-Scougall, I 1949, 'Pigeon netting--sport of Basques', National Geographic 96(3):405-16.

Casey, DA 1934, 'An uncommon type of stone implement from Australia and New Guinea', Memoirs of the National Museum of Victoria 8:94-9.

-- 1936, 'Ethnological notes', Memoirs of the National Museum of Victoria 9:90-7.

Clement, E 1903, 'Ethnographical notes on the Western-Australian Aborigines. With a descriptive catalogue of a collection of ethnographical objects from Western Australia by JDE Schmeltz', International Archiv for Ethnographie 16:1-29.

Eyre, EJ 1964 (1845), Journals of overland expeditions of discovery into central Australia and overland from Adelaide to King George's Sound, in the years 1840-1, T&W Boone, London. Libraries Board of South Australia, Adelaide (Australiana Facsimile Editions).

Golson, J 2001, 'New Guinea, Australia and the Sahul connection', in A Anderson, I Lilley & S O'Connor (eds), Histories of old ages: essays in honour of Rhys Jones, Pandanus Books, Research School of Pacific and Asian Studies, Australian National University, pp. 185-210.

Gregory, AC 1819-1905, Journals of Australian explorations, Government Printer, Brisbane.

Groube, L 1986, 'Waisted axes of Asia, Melanesia and Australia', in GK Ward (ed.), Archaeology at ANZAAS, Canberra, Canberra Archaeology Society, Canberra, pp. 168-77.

McCarthy, FD 1940, 'Comparison of the prehistory of Australia with that of Indo-China, the Malay Peninsula and the Netherlands East Indies', Proceedings of the 3rd Congress of Prehistorians of the Far East. Singapore 24/1-30/1/1938, Govt. Printing Office, Singapore, pp. 30-50.

-- 1941, 'Aboriginal ritual and mystery stones', Australian Museum Magazine 7(11):366-72.

-- 1944, 'Some unusual stone artefacts from Australia and New Guinea', Records of the Australian Museum 21(5):264-6.

-- 1952, 'Some new records of tanged implements and pounders in Eastern Australia', Mankind 4(9):361-4.

-- 1976, Australian Aboriginal stone implements, including bone, shell and teeth implements, Australian Museum, Sydney.

McCarthy, FD, Bramell, E & Noone, HVV 1946, The stone implements of Australia, Australian Museum, Sydney (Memoir IX).

McCourt, T 1975, Aboriginal artefacts, Rigby, Adelaide.

Peggs, AJ 1903, 'Notes on the Aborigines of Roebuck Bay, Western Australia', Folklore 14:324-67.

Petri-Odermann, G 1963, 'Das Meer im leben einer nordwestaustralischen kustenberoltkerung' (The sea in the life of a northwest coastal population), Paideuma 9:1-17.

Piddington, R & Piddington, M 1932, 'Report of fieldwork in northwestern Australia', Oceania 2(3):342-58.

Seligman, CG 1915, 'Note on an obsidian axe blade from Papua', Man 91:161-2.

Seligman, CG & Joyce, TA 1907, 'On prehistoric objects in British New Guinea', in WH Rivers, RR Marrett & NW Thomas (eds), Anthropological essays presented to Edward Burnett Tylor, Oxford University Press, pp. 325-41.

Smith, M & Kalotas, AC 1985, 'Bardi plants: an annotated list of plants and their use by the Bardi Aborigines of Dampierland, in north-western Australia', Records of the Western Australian Museum 12(3):317-59.

Thorpe, WW 1928, "Ethnological notes 1', Records of the Australian Museum 16(5):241-53.

Withnell, JG 1965 (1901), The customs and traditions of the Aboriginal natives of north western Australia, Libraries Board of South Australia, Adelaide (Australiana Facsimiles Editions).

Kim Akerman is an independent curator, consultant, archaeologist, anthropologist and art historian. He has worked in the Kimberley region since the 1960s, and has written extensively on the cultural materials of Indigenous Australians.

4 Dorset St, Moonah, Tas. 7009, <kim.akerman@bigpond.com>