Volume 74 1965 > Volume 74, No. 1 > Archaeological excavations in New Guinea: An interim report, by J. Peter White, p 40 - 56
ARCHAEOLOGICAL EXCAVATIONS IN NEW GUINEA: AN INTERIM REPORT 1
The prehistory of Papua-New Guinea is almost unknown. Collections of prehistoric mortars, animal figurines and axes, usually without recorded context or provenance, have been made since the late nineteenth century. Some excavations were made as early as 1905 2, but properly controlled and recorded data has become available only since 1959-60. Then, excavations in the Western and Central Highlands showed that a relatively unchanging flaked stone industry, including pebble tools and large utilised flakes, had existed for some 10,000 years 3. Polished axe-adze blades, “waisted blades” 4, and finally planilateral sectioned polished axe blades appeared in more recent times. By 1964 at least six archaeological questions could be profitably raised:
In the early part of 1964 an archaeological survey in areas of the coast, foothills and highlands of Papua-New Guinea suggested that in order to produce significant results a long period of survey would be required for the coast and foothill areas 8. This time was not available to me, and it was therefore decided to restrict work to the highlands areas, concentrating on the first four problems outlined above.
From June to November, 1964, surveys and excavations were made in two areas—part of the highlands of the Central District of Papua and the Eastern Highlands District of New Guinea.
This paper is an interim report of the work, prepared to outline its scope. It is stressed that little detailed work has yet been carried out on any of the material. All statements, and especially all interpretations, may be subject to revision as the work proceeds.
Kosipe, Central District, Papua.
In 1960 building operations produced a number of artefacts, including two stone bowls (mortars), axes and waisted blades. These were reported to be associated with carbon and hearths 9. It was hoped that this site might produce these artefacts in some datable cultural context.- 42 - 43
Kosipe Catholic Mission lies about 11 miles north-northwest of Woitape government station, at an altitude of 6,400 ft. (uncorr.). The mission is situated on a narrow ridge running down to a large swamp about 30 metres away and 10 metres below. Excavations over an area of 42 square metres were made at each end of the church, where the building operations had produced artefacts. Taken to an average depth of 1 metre, the dig produced a consistent stratigraphy of a buried soil with carbon and artefacts lying below about 30 cm. of sterile material 10. Six waisted blades, three axes, four flake tools and a small number of flakes were found in the buried soil. The carbon found was in the form of lumps, and no hearths, bones or other traces of occupation could be detected. Carbon samples have been submitted for age determination.
The flake artefacts from this site seemed to be unlike those found by Bulmer 11, but unfortunately were too few in number to allow significant statements to be made. However, the site may give us a date for the use of these artefacts, and also, through a study of pollen grains if these exist, allow us to say whether these artefacts belong to the period of sweet potato cultivation.
Eastern Highlands District, New Guinea.
I excavated two sites in the E.H.D. and also inspected some 85 caves and shelters. A number of these had remains of recent occupation, but only 7 had habitation deposit of a depth greater than 20 cm. The seven sites consist of two in the Lamari River valley south of Kainantu, two in the Asaro River valley, one at the north end and one near the junction with the Dunantina River, one near Chuave, on the backslope of Mt. Elimbari, and two in the Chimbu valley north of Kundiawa. One village site, abandoned shortly before European contact times, was found in the Asaro valley.
Numerous rock paintings, mostly in geometric forms, were recorded in the Lamari, Asaro and Chimbu valleys and in the area south of Chuave. Some semi-representational petroglyphs were found near the more southerly shelter site in the Asaro valley.
Aibura Cave, Lamari River Valley 12.
This site lies on the eastern edge of the highlands. As such it is well within the eastern sub-region and could provide a basis for testing the postulated cultural division between east and west. Locally, the site is within the area being studied by the University of Washington Micro-evolution Project 13. Their linguistic and anthropological studies suggest - 44 that the area has been culturally stable over some period of time, and I thought that I might be able to test this in archaeological terms.
The cave is located about 14 road miles SSE of Kainantu on the road to Obura Patrol Post. It lies about 25 minutes walk south-west of Barabuna No. 2 village. The altitude is 5,300ft (uncorr.). The cave is formed by the hollow interior of an isolated block of Tertiary limestone some 100 m. long, 30 m. wide and 15 m. high. This block now rises out of a small swampy valley which drains into a small west-flowing stream (Kondanauta R.) about 50 m. north of the cave. The valley floor is covered with pit-pit (Phragmites). The alignment of the cave is north-northwest. It consists of two parts, a main chamber which is lit by several small holes in the roof and by an entrance on the west side, and a long tunnel which is very dark. There are four other entrances to the cave, but all are low and slight bends prevent much light from entering. The lip of the west entrance is some 6 metres from the main chamber and there is a 2 m. high talus slope reaching from the entrance to the chamber. Outside this entrance the talus slope is very steep. The floor of the chamber is flat. The cave is now inhabited by some swiftlets, which are hunted by the local people from time to time.
An area of twelve square metres (4 x 3 m.) together with an extension of three square metres was excavated on the west side of the chamber, just to the north of the western entrance passage. Excavation was done as far as possible according to visible layers. All artefacts which were recognised as such during excavation were measured into position in three dimensions.
Occupational remains were found to a depth of 1.3 m. Below this was a water-deposited granular soil mixed with angular particles of limestone. Ground water from the pit-pit swamp outside was encountered at a depth of 2.1 m.
The upper 60 cm. of occupation consisted of many white ash lenses, which had been spread and compacted by water action, probably drips from the roof. Below this lenses were not visible, and have probably been removed by water running through the soil in the wet season. In the upper part of the deposit postholes were located. Many of these were in a series of lines most of which were set approximately parallel to the west wall of the chamber and some 2-3 m. from it. These are probably the remains of small wind-breaks, consisting of stakes with some interwoven vegetable matter. This type of shelter has been seen in the area when people make overnight stops at rock shelters 14. Villagers from Barabuna claimed that the cave had been used in pre-European times as a refuge during tribal fighting, and it may be that these shelters were used then.
1. Flaked Stone.
Approximately 600 artefacts and 500 flakes and chips of stone were found. From the low proportion of flake and chips it seems likely that this site was not a workshop, although some tool-making probably occurred. Artefacts were made mostly from cherts and quartzites, with occasional pieces of mudstone and obsidian.- 45
The most common form is a piece or thick flake with retouch or utilisation on one or more of the sharp edges. This secondary work may be termed step-flaking and it always occurs on one side of the edge only. the other side being cortex or unretouched flake surface. A number of criteria might be used to divide up these artefacts. For example, one may distinguish those with straight working edges, notched pieces, pieces with fine retouch, those with retouch on several edges and so on. However, in cases where metrical tests are applicable, as for example in distinguishing notched from straight-edged tools, these tests tend to suggest that ‘types’ which one can create are merely extremes of a particular range. It is quite clear that types in the Australian sense of readily distinguishable forms do not exist, and it may be that traditional methods of typological division are not adequate to describe this industry accurately.
Other flaked tools found include four hammerstones, a few cores and a few pieces where utilisation has removed small irregular flakes on both sides of the edge.
2. Polished Stone.
3. Bone Tools.
Eighteen small sherds were found, all from above 60 cm. depth. Four fragments are rim sherds. Pottery is not now made in the area, but is traded in from the upper Markham River valley 30 miles away and from the Agarabi people who live 12-15 miles to the northeast 20. Two sherds are similar to Markham pottery collected in the area in 1964. This is coarse in texture, thick and decorated with incised lines and gouges. At least ten sherds are much finer, both as to texture and thickness, than either the Markham or Agarabi pottery. Watson 21 gives the average thickness of 10 fragments of Agarabi pottery as five-eighths inch (16 mm.). The average thickness of eleven sherds from Aibura is only 5.3 mm. with a range from 2.8 to 9.0 mm.
Twenty-three fragments of sea shell and forty fragments of freshwater mussel were found, the majority in the upper levels. Dr. D. F. McMichael, Curator of Molluscs, Australian Museum, who very kindly studied the shells, gives the following identifications for the sea shells: Ovula ovum L. (the Egg Cowry), Trochus niloticus L. (the Commercial Trochus shell), Cypraea annulus L. (Ringed Money Cowry), Nassarius thersites Bruguiere (Plicated Dog Whelk), Charonia tritonis L. (Trumpet Shell) and Oliva sp. (Olive Shell) and Gymatium s.l. or Murex s.l. These identifications add at least two families, namely Charonia and Oliva to the list of sea shells found in the Highlands 22. The presence of two families, namely Ovula and Trochus, is confirmed.- 47
Considerable quantities of ochreous material (total weight 336 gm.) in various reds, yellows and creams were found, in all levels. One piece of it has a working groove. The modern equivalent is trade-store poster paint, a sample of which was found 1 cm. below the surface.
7. Human remains.
A number of fragments of human skeletal material were recovered. All are in a very fragmentary condition, and none were found in any condition even approximating to a burial. They are being studied by Dr. L. Freedman, Department of Anatomy, University of Sydney.
8. Food remains.
Animal bones were found in some quantity. The animals represented include at least two species of macropods (kangaroo/wallaby), cuscus, possum and smaller phalangers, several dasyurids, large and small rodents, snake (? python), birds and bats 23. Pieces of cassowary eggshell indicate that eggs were eaten, but few cassowary bones were present. Occasional pig and dog bones were present, which must come from domestic or feral animals.
Considering the importance of pig and dog in the present village economy, the number of bones of these animals in the site is surprisingly low. Possible explanations for this might include: (a) Pigs are not part of the everyday diet, but are eaten on ceremonial occasions. Under normal conditions such ceremonies are unlikely to take place in a cave. (b) If the site were a refuge and occupied only temporarily at other times it is unlikely that pigs would be killed there and it would be rare that a dog would die there.
Among the macropod and possum bones, mandibles are particularly common. It is normal practice in this area to keep mandibles of animals caught by hunting hung up in the huts, and this may account for the large numbers of these bones in the site. It is interesting to note that the majority of animals are probably of forest-living species. The immediate area around the site is now covered by kunai grassland and the lower montane rainforest 24 is 3-5 miles away. Nowadays hunting parties still go to the forest to hunt, but this is not very common, and the number of animals brought back is small, and they do not seem to play an important part in the diet. The archaeological evidence of considerable numbers of bones in the deposit would seem to run counter to current observation. It may be that more detailed studies will be useful in helping to pinpoint areas where archaeological inferences may lead one astray, or, alternatively, suggest where current anthropological information is not a good guide to prehistoric practices.- 48
The interior of the cave is covered by more than 100 paintings. These are made in two techniques—lines of white dots forming a design, and white smears, often outlined with charcoal 25. While excavation was in progress an old man arrived uninvited and made a painting in the cave, using the white dot technique. The design created was a complex pattern of circles and lines, but if there was a meaning to it he would not reveal it. The material used for painting was fine white ash obtained from burning pit-pit stems and leaves. The ash, from which all lumps of carbon have been removed, was moulded lightly in the left hand. A pinch of it was taken and pressed with the right thumb onto the rough, soot-blackened cave wall. Close inspection of other white dot paintings showed that the same raw material had been used there. On some dots thumb prints could still be seen.
On two separate occasions, six days apart, this man was asked to identify the paintings already in the cave. On the first occasion he gave names to thirty out of thirty-four designs. The range of objects included plants, animals and birds, the sun, sky and moon, caps and headresses used on ceremonial occasions, men and spirits. Later he re-identified 20 designs. Of these, four were the same as the previous identification, and these included three which were obviously non-representational.
It is probable that there is considerably more ethnographic and anthropological material of interest to archaeologists to be gained from a detailed study of the painters and painting.
Niobe Shelter, near Chuave.
Excavation of this shelter was undertaken for three main reasons. It was hoped that mortars, pestles, figurines and club-heads, which occur in this area, might be found in a more detailed context. Since we know that the area has been occupied for at least 10,000 years (which we know for nowhere else in New Guinea) I thought that there were high possibilities of a discovery of this kind. Also, I thought that it might be useful to try to carry up to the present the Kiowa sequence, which, because of disturbances and abandonment, ceases to be of definite value at about 4,000 B.P. 26. It was also hoped to collect pollen and faunal evidence for the relatively recent prehistory of the area, and possibly for the earlier periods.
The shelter is situated on the backslope of Mt. Elimbari, about 3 miles south-southwest of Chuave and about 2 miles from the Kiowa site. Its altitude is 5,400 ft. (uncorr.). Niobe shelter is formed by a limestone overhang and is probably the minor entrance to a major cave system. The actual sheltered area visible is some 18 metres long and 4 metres wide, and is aligned in a north-northwesterly direction. The greatest breadth of the shelter is about six metres, where the overhang meets the cave entrance. Along the outer side of the shelter, though still under cover, lies a heavily used foot-track. In times of rain it is common for people - 49 to light a small fire under the shelter and perhaps cook a little food. Work was considerably hampered by this and by the continual passage of interested spectators.
Initially it proved difficult to get permission to excavate the site. No less than three different ‘lines’ owned a part of it and were not amenable to its being disturbed. Because of this a trench two metres by 1.5 metres aligned approximately east-west was dug rather rapidly on the south side of the cave entrance. When the site proved productive, month-long negotiations took place and eventually I gained permission to excavate further, though only in a certain part of the site.
The major excavation consisted of a trench 5 x 1 metres (A3-A7) reaching from outside the dripline (A7) to a point in the cave entrance where the headroom was about 50 cm. (A3). This was dug along the north wall of the test trench. A further four square metres in the form of (a) a 2 x 1 m. extension eastwards of the test trench (Z6-Z7), and (b) a 2 x 1 m. cutting on the north wall of A6-A7 (B6-B7), was also excavated. Both these cuttings ran across the broad dripline. The total area of the main excavation was therefore 9 sq. metres.
Towards the back of the shelter large blocks of the deposit were heavily concreted with lime. Much of this concreted soil was capped by slabs (4-10 cm. thick) of sterile stalagmite, which lay 20-30 cm. below ground surface. The stalagmite slabs sloped downwards towards the dripline and ended abruptly in A6 and B6 about 50 cm. inside the inside edge of the dripline. They ended about two metres from their highest point, which lay in A4. Although clearly formed in situ as a single block, the upper part of this concreted material was now cracked into several pieces. Cultural material in a soft dry matrix was found in the fissures between and slightly underneath the edges of the blocks, to a depth of about 1 metre below the ground surface. The concreted soil contained large amounts of bone, but facilities were not available to extract this.
In front of the concreted material the soil was a uniform dark red-brown, becoming more clayey under and in front of the dripline. It contained many small blocks of limestone from the roof, and these were concentrated above about 75 cm. below the surface. At the base of the excavation two very large limestone boulders, very rounded by weathering during their period of exposure prior to the deposit building up around them, were encountered. In this area the site was extremely rich in stone artefacts and bone material. It was excavated one square metre at a time in spits of 5 to 8 cm. All artefacts were measured in three dimensions. The spits were sloped slightly in conformity with the surface slope, and nine levels were taken at the base of each spit. This should give sufficient control for analytical purposes.
Visible layers were scarcely apparent in the whole depth excavated and artefacts cannot be grouped according to them. Further, there are good reasons for doubting that depth within the deposit will provide even a relative chronology of the artefacts.
It can be said definitely that any material within the stalagmitic soil is older than any material in front or on top of it. For the rest of the - 50 deposit distinctions based on depth are probably to be regarded as dubious. Firstly, although slight changes in soil colour could be seen in the sections, these changes appear to be related to the dripline and other natural phenomena rather than to human activity. Then there is the evidence of the artefacts. As with the Aibura material the finds are generally so undifferentiated as to make stratigraphic divisions on the basis of typology implausible. The correlate of this is that it is difficult to decide whether, and if so how much, the deposit has been disturbed. There is one piece of evidence which suggests that natural disturbance has been considerable. This consists of three fragments of the same polished greywacke axe.
The fragments are a butt, which is pointed, a section of one side, and a part of the blade, which is flaked along the cutting edge. In cross section the axe is lenticular with squared sides, each side exhibiting three bevels. The length of the axe is 11.9 cm., the maximum width measurable 4.8 cm. and the thickness 2.1 cm.
The position of these three fragments is as follows: the butt, which is the largest piece (wgt. 84.5 gms.), was found in A3, the innermost square, at a depth of 2 cm. below the surface. The other two fragments were found 3.95 and 3.99 metres towards the front of the shelter, from the butt fragment (junction A7/B7) 70 cm. towards the north and at depths of 69 cm and 62 cm. below surface respectively.
This evidence can be taken as documenting the effect of one or more of several processes, acting on archaeological deposits 27.
With the possibility of these various factors operating it would clearly be incorrect to claim that precise assemblages can be defined in terms of a particular stratigraphic level. It may be that more of the broken artefacts or flakes can be fitted together thus allowing a more general picture of disturbance to emerge. Only then could claims to a relative chronology for the deposit be upheld.
Taking a wider view, it might be pointed out that an extremely common method of dating a deposit or level is by the most recent element within it. This is valid only if it can be shown that the material used for dating genuinely belongs to that position. In the absence of ‘type fossils’ the concept of proportion may be used. If all but one or two of a particular kind of artefact come from a certain level then it would be unwise to overstress the exception. However, this concept is not yet applicable to New Guinea where the chorological and chronological limits of artefacts and industries have not yet been defined. These factors then must affect the interpretation both of the artefactual material and of the ecological data.
1. Flaked stone.
About 800 flaked stone artefacts were found, mostly made in cherts, quartzites and basaltic rocks. Primary analysis only has been done as yet, and numbers are approximations only, but the following groups are suggested:
These can be divided into two distinct classes.
(a) Very thin rectangular blades (5 fragments). The most complete of these blades is 10.4 cm. long, 5.8 cm. wide and a maximum of 1 cm. thick. It is ground all over and has a very flat lenticular cross section. The sides are parallel, and the blade set at right angles to the sides. The other four fragments are from 0.4 to 0.9 cm. thick. Two exhibit the straight-sidedness of the large fragment and one of these shows a right-angled corner section. This latter fragment shows one edge (possibly the blade?) with symmetrical bevelling. The other two fragments are very thin and show slightly curved edges.
The largest fragment is made of quartzite, the other four of shale (?). All five artefacts were found within 50 cm. of the surface. Although there is no evidence yet on which to date any part of this site it is tempting (if far-fetched) to consider these blades as of recent manufacture, and possibly skeuomorphs on some kind of European artefact (bush knives?). They do not appear to fit into the normal pattern of axe typology 31. One artefact of this kind is known from the Yodda goldfields near Bulolo 32.
(b) Other axes and fragments: These comprise three whole axes and 15 fragments. Of the whole axes one is lenticular in cross section with squared sides, the other two are lenticular. The first is ground nearly all over, has a slightly rounded butt, a curved blade set slightly asymmetrically when viewed from the side and is made in light green greywacke. Of the other two, one has a squared butt, and an asymmetrically curved blade set asymmetrically to the long axis. The other is flaked all over and is symmetrical with a semi-squared blade and butt.
Three of the axe fragments have been described earlier (p. 50). The other twelve comprise two butts one very square, from partly polished axes of lenticular cross-section; six fragments of polished and three of flaked axes and one axe roughout. The roughout is a chunk of river pebble bifacially flaked along one side in a method characteristic of other flake axes. The other edge has been used as a pebble tool. The pebble is broken laterally. This fragment documents the local manufacture of some axes 33.- 53
(c) Waisted blade: One waisted blade was found at a depth of 81 cm. below the surface. It is shouldered rather than waisted. It measures 7.7 x 5.2 x 2.2 cm. The blade is ground on both sides but the shoulder, which commences 3.1 cm. from the blade, is flaked. The artefact is comparable in size to those excavated by Bulmer 34.
One fragment of mortar was excavated in A3, 37 cm. below the surface. It is a wedge-shaped segment of a plain stone dish, made in tuffaceous stone. Both outer and inner surfaces form a smooth curve and there is no sign of a flattened base. The rim appears to be nearly circular; if it were so it would have a circumference of about 57 cm. (radius 9.1 cm.). It would appear to stand about 5.6 cm. high. Apart from the rim, the thickness of the stone ranges from 1.25 to 1.52 cm.
A fragment of stone rim (5.9 x 1.2 cm.) was found at a depth of 31 cm. below the surface in A7. It also is of tuffaceous stone and is very similar in shape to the rim of the larger fragment. This would seem to place mortars into a fairly recent context, and it also suggests that they are part of the normal industrial activity of this area.
4. Club Head.
A piece of triangular-sectioned stone club head was found in B7, 48 cm. below the surface. The hole for the handle has been worked from both sides, and the circumference of this hole at each end is 12.5 and 11.9 cm. The segment stands 5.5 cms. high. It is clear from the segment found that the club was not symmetrical about its handle, but must have been considerably off balance.
5. Bone Tools.
No detailed study has been made of these, but at least 4 types are suggested.
Shell fragments, which have been studied by Dr. D. F. McMichael, are mostly of freshwater mussels. Sea shells include 2 fragments of Oliva and a piece of Cypraea, probably annulus L.
7. Food Remains.
About 100 lbs. weight of bone, much of it in a very broken condition, was excavated. Such a quantity of bone will enable a detailed study of the range of animals hunted, even if at this site we can say little about changes in hunting patterns over time. One point to emerge already is that eggshell is quite common at this site.
It is interesting to note that when Europeans first entered the area hunting was not an important part of everyday life. Salisbury 37, discussing indigenous economic life, says: “Against (the) picture of diverse vegetable resources must be set the scarcity of animals. Hunting is of little importance for the supply of food, though both birds and opossums are shot for sport and incidentally for food. . . . Smaller game, rats, grubs and insects are often caught by youths, but are eaten only by women and children.” Such a picture is not entirely compatible with the quantity of bone recovered from Niobe. It may be that the accumulation of bone at the site took place over a longer period of time than I am inclined at the moment to think, or that hunting only recently became of so little importance, or that for some reason hunted animals were eaten in more than usual numbers at this site. Alternatively, it might be suggested that the recent development in sweet potato cultivation led to a decrease in hunting activity. But I have no good grounds for thinking that there has been any noticeable change in the amount of hunting recorded at this site, which has been in use right up to the present.
The most interesting fact to emerge from the work so far is that there appears to be some archaeological confirmation of the cultural break between the eastern fringes and the rest of the highlands. This is seen particularly in terms of the pebble tools which have been found in the highlands from Baiyer River through Chuave to the Asaro River valley 38. They were totally absent from the Aibura site in the east although river pebbles were available at no great distance away. I hope to check this theory by excavating at another site on the Lamari River valley.
It also seems that, when contrasted with the rest of the highlands, the eastern fringes have much more contact with the lowlands of the Markham River valley in terms of trade in pottery and shell. This is to be expected on geographic grounds alone, but the contrast is rather more marked than I had expected.- 55
The possibilities of ecological studies are considerable. The large quantities of bone preserved at the sites should give us some idea of how long domestic dog, pig and hen have been in the highlands. It is possible too that the study of pollen grains will provide a more definite idea of the time of arrival of the sweet potato and its impact on the material culture of the area 39.
The use of ethnological and anthropological studies in the interpretation of archaeological material seems to offer considerable promise. I have suggested in this paper some ways in which this might be done, but as more archaeological work is done more detailed and more sophisticated analyses will become possible.
Most importantly, perhaps, there is in New Guinea the possibility of observing men who can still make stone tools. Some work in this field has already been done among the Kukukuku 40, and the making of Hagen axes has been recorded 41. During 1965 I am proposing to work among one of the very few groups of people in the Asaro valley who can still make pebble tools. Most of the people in this area have already forgotten these techniques, but there are considerable areas in the Western highlands where European contact is so recent and so superficial that stone toolmaking techniques are preserved until today. If there is one urgent field for archaeological research in New Guinea, this, the making and use of stone tools, is it.
1 The field work reported here was financed by the Department of Anthropology and Sociology, Australian National University. I am grateful to Professor J. A. Barnes and Mr. J. Golson for their encouragement. I would also like to thank the Director of the Department of District Affairs, Mr. J. K. McCarthy, Government Anthropologist Mr. C. Julius, and the many people who helped me during this work, especially D. Cole, Fr. M. Gasser, F. Parker, P. J. Thomas, W. E. Tomasetti, P. Wohlers, and Professor J. Watson.
2 Pöch 1907.
3 Bulmer 1964a, 1964b.
4 Bulmer 1964a:257.
5 Read 1954:12.
6 Watson 1964:3; Bulmer 1964c.
7 Mulvaney 1964.
8 White 1965.
9 Information kindly provided by W. E. Tomasetti, then Assistant District Officer, Tapini, on the basis of information from Kosipe Mission.
10 I wish to thank Dr. K. Cook, Dept. of Geology, School of General Studies, Australian National University, for his help in analysing this and other soils.
11 I am grateful to Mrs. S. Bulmer for permission to look over her material and for discussing it with me. The material is now housed in the Dept. of Anthropology, University of Auckland.
12 The cave was discovered by Mr. A. Vincent of the Summer Institute of Linguistics in 1961 and was drawn to my attention in 1964 by Dr. V. Watson of Seattle University.
13 Watson 1963.
14 P. J. Thomas, Obura Patrol Post: personal communication.
15 Cf. Bulmer 1964c:69.
16 Blackwood 1950.
17 Personal observation.
18 Bulmer 1964c:55.
19 Willcox 1963:39.
20 Watson 1955:127.
21 Watson 1955:123.
22 Bulmer 1964c:56.
23 I wish to thank Mr. B. J. Marlow of the Australian Museum, Sydney, Dr. J. Calaby, Wildlife Survey Section, Commonwealth Scientific and Industrial Research Organisation, and Mr. R. Barwick, Dept. of Zoology, School of General Studies, A.N.U., for help with the identification of animal remains.
24 Brookfield 1964:23.
25 White 1964.
26 Bulmer 1964c:59.
27 I am grateful to Mr. J. M. Matthews and Mrs. C. White for discussing these problems with me.
28 Mulvaney 1961:67.
29 I am indebted to Mr. F. D. McCarthy, Director of the Australian Institute of Aboriginal Studies for this suggestion.
30 Semenov 1964.
31 Bulmer 1964a: 246-55.
32 Information kindly provided by Mr. F. D. McCarthy.
33 Bulmer 1964a:252.
34 Bulmer 1964a:262.
35 Mulvaney 1961:84.
36 Bulmer 1964c:60.
37 Salisbury 1962:44.
38 Bulmer 1964a; D. Cole:personal communication.
39 Buller 1964c:45-8.
40 Blackwood 1950.
41 Vial 1940.