Excavations, Surveys and Heritage Management in Victoria

Volume 3

2014 Excavations, Surveys and Heritage Management in Victoria

Volume 3

2014

Excavations, Surveys and Heritage Management in Victoria Voume 3, 2014

Edited by

Caroline Spry David Frankel Susan Lawrence Ilya Berelov Shaun Canning

Front cover photograph Sisters Rocks. Courtesy Darren Griffin (photo taken on 17 October 2013)

2 Contents

Editorial note 5

Papers

Pleistocene to Holocene in the Upper Maribyrnong River Basin: excavations at Deep Creek, 7 Bulla William Anderson and Racheal Minos

Investigation of a Pleistocene river terrace at Birrarrung Park, Lower Templestowe, Victoria 15

Martin Lawler, Ilya Berelov and Tim Cavanagh

A Pleistocene date at Chelsea Heights, Victoria: evidence for Aboriginal occupation beneath 23 the Carrum Swamp Jim Wheeler, Alan N. Williams, Stacey Kennedy, Phillip S. Toms and Peter Mitchell

Results of recent archaeological investigations along Wallpolla Creek, northwestern Victoria, 33 Australia Ben Watson and Paul Kucera

The Browns Creek Community Archaeology Project: preliminary results from the survey and 43 excavation of a late Holocene shell midden on the Victorian coast Martin Lawler, Ron Arnold, Kasey F. Robb, Andy I.R. Herries, Tya Lovett, Christine Keogh, Matthew Phelan, Steven E. Falconer, Patricia L. Fall, Tiffany James-Lee and Ilya Berelov Hiding in plain sight: excavation of a lost pioneer homestead on Kororoit Creek 53 Barry Green

Lonsdale Street, Little Lonsdale Street and Altson Lane, Melbourne: initial testing and first 57 stage of historical excavations in 2013 Alison O’Connor

Socio-economics of archaeology in Victoria after the Aboriginal Heritage Act 2006 65

Nicolas Zorzin

3 Contents

Abstracts

Identifying Aboriginal hearths in Late Quaternary northwest Victoria: cultural vs natural 75 accumulations of burnt clay and the implications for understanding the archaeological record Alex Blackwood, Emily Dillon, Chris Silvester, David Clark, Jeffrey Clarke, Maurizio Campanelli, Anthony Dall’Oste, Ada Dinckal, Sara Lombardo, Anthony Romano, Adam Valka, and Jillian Garvey Investigating Late Quaternary human occupation of semi-arid northwest Victoria: preliminary 76 survey and excavation at Neds Corner Station and Murray-Sunset National Park Jillian Garvey

Re-evaluating artefact density as a standalone method of categorising stone artefact 77 occurrences in Victoria: a case study from Beveridge Caroline Spry

Landscape analysis in archaeology: interpretation of Aboriginal campsites at ‘Greenhills’, 78 Pakenham David Rhodes

The Victorian Digital Archaeological Data Archive: an annual update 79

Josara de Lange

Indigenous cultural heritage management at the Mornington Peninsula Shire 80 Adam Magennis

Sisters Rocks: changing connections to a sacred place 81

Darren Griffin and Ben Watson

The formation of cesspit deposits at Little Lon, Melbourne 82

Sarah Hayes and Barbara Minchinton

The buildings of Abbotsford Convent: insights into the practice of confinement 83

Edwina Kay

Two households… in Truganina, where we lay our scene 84

Sylvana Szydzik and Pamela Ricardi

A brief comparison of Indigenous cultural heritage legislation - or - It’s not that bad here in 85 Victoria….no, really…. Jamin Moon

La Trobe University’s Young Archaeologists’ Program 86 Thea Kinsela, Maddy Maitri and Susan Lawrence

Fieldwork at Keilor: a short film 87 John Tunn

4 Editorial note

The papers included here were presented at the third of work being carried out in Victoria, and we hope that colloquium on Victorian archaeology held at La Trobe many of these will also form the basis for more complete University in February 2014. As in previous years, we studies in the future. The extended papers were refereed, would like to thank all the participants whose attendance and we would like to thank those who assisted with this testifies to the importance of what has now become task. Caroline Spry managed this process and much of a fixture within the local archaeological calendar as the sub-editing of this volume. Layout, as in other years, an important opportunity for academics, consultants, was undertaken by David Frankel. managers and Aboriginal community groups to share As with the previous volumes, the illustrations are their common interests in the archaeology and heritage all in black and white, although many, especially the GIS of Victoria. maps, would be clearer in their original colour. Anyone It is of particular value to include papers in this issue interested in having copies of these should contact the on the earliest periods which are far less well-represented individual authors. in Victoria than in many other parts of Australia. That The 2014 Colloquium was generously supported these data emerge from consultancies demonstrates the by our Principal Sponsor, Australian Cultural Heritage importance of a strong research component in heritage Management (ACHM), as well as other sponsors, studies. But many othjer aspects of both research and including Ochre Imprints, AACAI, Biosis Pty Ltd, community engagement are also covered, dealing both Archaeological and Heritage Management Solutions with Aboriginal and European heritage. (AHMS) and the Department of Archaeology at La We are especially grateful to those who later were Trobe University. able to prepare more developed papers for publication. The editors and authors acknowledge the traditional Other presenters allowed slides abstracted from their owners of the land and heritage discussed in this book, PowerPoint presentations to be included with their and pay their respects to their elders, past and present. abstracts in this volume. These demonstrate the range

5 6 Pleistocene to Holocene in the Upper Maribyrnong River Basin: excavations at Deep Creek, Bulla

William Anderson and Racheal Minosa

Abstract Geomorphological research, supported by radiocarbon age estimates from artefact-bearing Archaeological study of the Maribyrnong River valley north sediments, revealed a long sequence of human of Melbourne has been crucial to understanding long-term occupation at Keilor dating from the late Pleistocene and human habitation in Australia. A recent cultural heritage early Holocene (Munro 1998). The earliest artefacts were investigation across the incised valley of Deep Creek at contained in clay to depths of 7 m from the so-called Bulla has yielded new information on the chronology and Arundel Terrace, with age estimates greater than 31,000 context of Aboriginal occupation along this tributary of the BP, alongside megafaunal remains (Duncan 2001). Maribyrnong. Extensive lithic artefact scatters, featuring Another discovery of human remains in 1965 at Green diverse materials and forms, span a large alluvial terrace Gully (Brimbank Park), downstream and little more than beside a meander bend at the base of the valley. As well 3 km from the Keilor site, prompted a separate excavation as documenting the abundant artefacts present on the project that yielded substantial information on the surface and at shallow depths below ground, a mechanical archaeology and landscape history of the Maribyrnong excavation clarified the geomorphology of the terrace and River valley (Bowler 1970; Mulvaney 1970). Recent verified the presence of deeply buried cultural deposits. investigations at Brimbank Park have since built upon Quartzite implements were found within stratified silt the results of these earlier campaigns (Canning et al. deposits of up to 3 m depth, and charcoal samples were 2010; Tunn 1998). recovered from secure cultural contexts. Radiocarbon age Fieldwork and research at Keilor between the 1960s estimates obtained from these samples confirm human and 1980s were significant in shaping many of the occupation of this valley before and during the Pleistocene– practices, theories and debates still current in Australian Holocene transition. archaeology. However, the circumstances differed markedly from archaeological practice in Victoria today. The rise of cultural heritage management and Background the consolidation of privatised contract archaeology have been paralleled by a shift in emphasis towards The chance discovery in 1940 of a prehistoric human the investigation of Aboriginal places at risk of harm. skull at Keilor, north of Melbourne, signalled the start Research aims are fitted around pre-determined project of sustained archaeological interest in the Maribyrnong briefs — location, scale and budget — which are set by River valley. Recovered from eroded Pleistocene deposits, the sponsor rather than themselves being the motivating the Keilor cranium proved the antiquity of humans in factor. Thus, in the current context, the generation of new the region, and at the time it was judged to represent information and insights about the Aboriginal past are the oldest-known human skeletal remains in Australia seen as virtually incidental, and secondary to the aims (Gill 1966). Areas near the find-spot were investigated by of heritage management. Yet important information can Alexander Gallus, who sought to prove the development still be gleaned from commercial heritage assessments. of early hominins in Australia using an elaborate typology The opportunity to investigate Deep Creek, which has of stone tools (Mulvaney 1998). In part motivated by the many similarities to Keilor in terms of its topography and need to scrutinise these controversial claims, excavations geomorphology, is therefore particularly welcome. were undertaken by the Victoria Archaeological Survey and La Trobe University from 1977 to 1982 (Munro 1998:19–20). Investigations at Bulla a Dr Vincent Clark & Associates, 11/240 Sydney Road, Coburg, There are currently plans to construct a road bypass of Victoria 3058, Australia Bulla, a settlement 24 km northwest of central Melbourne [email protected] and 9 km north of Keilor (Figure 1). This will include +61 (03) 9386 4770

7 William Anderson and Racheal Minos building a modern viaduct which will replace the narrow Until recently, however, there has been little information and treacherous road that winds through Deep Creek to prove the chronology or nature of the Aboriginal valley, across a bluestone bridge built in 1869. Planning presence. The current investigation’s approach was for the Bulla bypass necessitated a Cultural Heritage informed by previous research, although the methodology Management Plan (CHMP number 11935), which could only be developed after having encountered the VicRoads commissioned Dr Vincent Clark & Associates physical and archaeological conditions. Large exposures to prepare in collaboration with the Wurundjeri Tribe of surface artefacts required a strategy that would record Land and Compensation Cultural Heritage Council, the their location and account for their attributes, extent area’s Registered Aboriginal Party. and density. It was important to establish the nature The CHMP addresses a large activity area between and context of this material and to explore the potential Moonee Ponds Creek in the east and the Sunbury plain for deeply buried deposits in similar situations to those to the west. Ground survey and test excavations carried found downstream, at Keilor. Two areas of particular out between late 2011 and early 2013 detected numerous interest were identified: on the granitic spur above the artefact scatters and resulted in 49 surface Aboriginal east of Deep Creek, where large, dense and diverse places being newly registered, some of which were later artefact scatters were recorded; and in the valley floor, amalgamated. By far the greatest concentrations of across alluvial terraces on both sides of the creek. artefacts recorded are near Deep Creek, on the summit and slopes of the tall spur east of the creek and the alluvial Bulla 1 terraces of the valley floor. Previous surveys have recorded Aboriginal places Extensive lithic artefact scatters that span an alluvial upstream from Bulla, along Emu Creek (Tulloch 2003) terrace to the east of Deep Creek, across an area of at least and Deep Creek near Wildwood and Mickleham (Rhodes 3,600 m², were recorded with the Victorian Aboriginal 2000; Richards and Sutherland 1995). Rockshelters with Heritage Register as Bulla 1 (VAHR 7822-3278). This associated artefact scatters were located during a recent Aboriginal place is located at a point where the creek survey along Deep Creek to the south of Bulla (Crocker meanders in a wide arc to the west, forming a sub-circular and Wheeler 2013). Along Jacksons Creek, which flows area divided in the centre by a low spur (Figure 2). The parallel to the west of Deep Creek, there are dozens of terrace is bounded to the north, west and south by Deep recorded Aboriginal places including artefact scatters, Creek, and by the valley slope to the east. The ground falls scarred trees and the earthworks known as the Sunbury sharply down to the creek, which is bordered by huge Rings which are thought to have been ceremonial granite boulders. At the foot of the valley sides are kaolin grounds, perhaps used for initiation rites (Frankel 1982; deposits and an abandoned quarry which operated from Sutherland and Richards 1994). On the plain east of before the 1860s (Anderson et al. 2013:12). Bulla, lithic artefact scatters and scarred trees have been More than 200 lithic artefacts were recorded on the recorded at Gellibrand Hill, east of Moonee Ponds Creek surface, though many more are present in the area. Raw (Presland 1983), and in the grey box forest around Radar materials include silcrete (60%), quartzite (18%), quartz Hill and the former village of Oaklands Junction, now (18%) and other materials (basalt, tachylyte, chert and within the perimeter of Melbourne Airport (Vines 1995). hornfels). Some nineteenth century items were also Whereas archaeological studies around Bulla only found, perhaps dating from the 1850s and 1860s when began in the late-twentieth century, interest in the area’s there was intensive agricultural and industrial use of the geology has been longer running. The incised valley area, including a large flour mill to the north (Lochton with its meanders and overlapping spurs was presented Mill, Victorian Heritage Inventory number H7822- by John Walter Gregory (1903:152) in his book The 2306). The density of surface artefacts was measured Geography of Victoria, and the area’s geological diversity systematically by documenting artefact numbers and was the subject of detailed analysis from an early stage types within sample blocks of 5 x 5 m, placed at different (James 1920). The valley bisects and exposes strata points across the area. These had an average density formed of Upper Ordovician and Lower Silurian shale of 24 artefacts per 100 m², the majority of items being and greywhacke, granitic rocks including the unique manufactured on silcrete (73%). The extensive and in Bulla Granodiorite, overlaid by lava flows of the so-called places dense surface scatters, the presence of diverse Newer Volcanics (Rossiter 2003:221); in the valley floor materials and forms including a large number of cores (7% are substantial alluvial and colluvial deposits. Apart from of the assemblage) and other evidence for primary and offering a textbook example of geological formations, secondary stone tool production such as two unfinished this diversity brought extractive industries from the mid- basalt axe blanks, indicate this to be a significant focus of nineteenth century, including quarrying of basalt and social and economic activity. kaolin (Moloney 1998). Seven pits measuring 1 m² each were excavated Deep Creek’s rich and diverse environment appears by hand at different points within the area of Bulla 1. to be the setting for millennia of human occupation. Lithic artefacts were found in all but one of these, at an

8 Pleistocene to Holocene in the Upper Maribyrnong River Basin: Excavations at Deep Creek, Bulla

Figure 1. Map of waterways northwest of Melbourne and Aboriginal places mentioned in the text (map drawn by Paul Kucera)

9 William Anderson and Racheal Minos

Figure 2. View of the terrace east of Deep Creek, the location of Bulla 1, from the hillside on the west side of the valley average rate of 24 items per 1 m² pit, mostly at shallow significant cultural deposits because of the considerable depths (< 200 mm) and contained in silt overlying rocky depth needed for the construction of bridge piers. This colluvium. Different conditions were encountered in led to further subsurface testing in the form of combined two pits positioned at the edge of the terrace, closer to mechanical and manual excavation. the creek bank, where there were deep and compacted deposits of dark brown clayey silt in which modern items Mechanical excavation were mixed with lithic artefacts. The presence of glass items at depths of up to 500 mm below the surface shows The principal aim of the mechanical excavation was to that the soil has been deposited in recent times, probably investigate the presence of cultural deposits at depths that as the result of flooding. The highly compacted soil made were difficult to reach using hand excavation alone. The excavation difficult, and although lithic artefacts were work took place over two days in February 2013. Using present, it was uncertain whether there would be any an excavator with a one metre-wide mud bucket, a trench undisturbed contexts at lower depths which would justify measuring 3 x 1 m was positioned close to the edge of the further arduous and time-consuming hand excavation. terrace, approximately 30 m from Deep Creek. Soil was The results of test excavations summarised above extracted and artefacts collected in regular intervals of demonstrate the subsurface aspect of cultural deposits 200 mm; all of the extracted soil was mechanically sieved at Bulla 1. However, they also highlight the insecure through 5 mm mesh. nature of contexts in which artefacts occur, at least to Sediment in the upper level consisted of dry, a depth of half a metre. To investigate the possibility of compacted, dark greyish-brown clayey silt which was intact cultural contexts, it was necessary to achieve a slightly acidic (pH 5.5) and contained plentiful and greater depth than was possible using hand excavation, diverse flaked lithic artefacts and one ground basalt axe within the project’s scope and budget. The question was head. There were also some nineteenth and twentieth important for the understanding and management of this century items, including a medallion dating from Aboriginal place and the possible avoidance of damage to 1911, commemorating the coronation of King George

10 Pleistocene to Holocene in the Upper Maribyrnong River Basin: Excavations at Deep Creek, Bulla

V and Queen Mary, at 0.8–1.0 m depth. At a depth of approximately 1 m, there was a distinctive change in the soil to fine-grained, reddish-brown silt with no clay content, which is alkaline (pH 8.5) (Figure 3). Aboriginal artefacts in this stratum were less frequent, but display more consistency in their materials, which are predominantly yellow and red quartzite, and feature different, larger forms, including a large horsehoof core, at 1.0–1.2 m depth (Figure 4). The artefacts present, the lack of intrusive modern items and the more homogenous silt are suggestive of a stratified and undisturbed cultural layer. Figure 4: Quartzite artefacts from the mechanically excavated pit Once the changed soil conditions were established, at Bulla 1: a horsehoof core from 1.2 m depth (left) and a scraper hand excavation was carried out to confirm the presence from 2.7 m depth (right) of artefacts in situ and to recover any dateable material. Both of these aims were achieved: artefacts were located silt that is also alkaline (pH 9), and initially contained within the silt, and two separate samples of charcoal were no artefacts (Context 3). However, a small number of obtained. Both samples comprised discrete, consolidated artefacts were recovered from depths between 2.7–3.0 clumps of carbonised material, and were identified at m. These include a large scraper of light and dark grey depths between 1.3–1.4 m, approximately 1 m apart from banded quartzite (Figure 5). Also from this depth are each other horizontally, and in the immediate vicinity of cylindrical and nodular concretions, first thought to flaked lithic artefacts (Figure 4). The samples are both be calcified plant roots, or rhizoliths, but which may from the centre of the upper silt stratum (Context 2). also comprise infilled worm tubes, or biotic casts, as Having verified the context of lithic artefacts and identified within the Keilor Terrace sediments (Bowler obtained charcoal samples, mechanical excavation 1970:22–23). Excavation ceased at 3 m, at which point resumed. Artefacts continued to be present up to 1.7 m, the machine was unable to dig any deeper without though the number dwindled. From below this depth, expanding the size of the trench. the soil comprises a slightly moist, fine-grained, yellow The three identified soil contexts each contain cultural material. Compacted sediment in the upper metre has been deposited recently, which is confirmed by the presence of the 1911 medallion and supported by the heterogeneous nature of lithic artefacts. The underlying silt strata are thought to be ancient and undisturbed deposits which have become sealed beneath the recent flood deposits. The largest number of lithic artefacts are from the upper 1 m (n = 166; 69%). Although the context is redeposited, there are some patterns in the material present at certain depths: a concentration

Figure 3. East wall of the excavation trench, showing three strata: (1) very dark brown clayey alluvium; (2) fine, reddish-brown silt; Figure 5: Charcoal sample (MEP-2) collected from the east wall of and (3) very fine, yellowish-brown silt the trench at 1.4 m depth

11 William Anderson and Racheal Minos of silcrete between 0.2–0.4 m, and a concentration of quartz, otherwise rare, between 0.6–1 m. A total of 70 lithic artefacts come from depths greater than 1 m, the majority being quartzite (63%), which in the upper 1 m constitutes only 14% of collected artefacts. There is also a difference in the size of artefacts, which in the upper 1 m have an average maximum dimension of 22 mm, but below 1 m have an average maximum dimension of 26 mm. These differences support the interpretation that underlying silt deposits and associated material culture belong to a distinct phase of occupation, in contrast to that represented in the topsoil and on the surface. The collected charcoal samples were submitted for radiocarbon (14C) dating to the University of Waikato Radiocarbon Dating Laboratory, New Zealand. The samples (MEP-1 and MEP-2) were cleaned and subjected to Accelerator Mass Spectrometry (AMS) measurement. Results of the analysis produced age estimates of 10,262 ± 27 BP and 10,086 ± 27 BP, calibrated to ranges of 12,142– 11,835 BP and 11,818–11,404 BP respectively (Figure 6). The general correspondence in the age estimates of samples collected at different positions in the trench, at depths of 100 mm apart and in close proximity to lithic artefacts, raises confidence in the association of the two charcoal samples with the deposition of the silt and the artefacts. In other words, the age estimates provide reliable information about the age of the silt deposits and the material culture contained therein. Figure 6. Results of AMS radiocarbon dating of two charcoal samples from Bulla 1 (analysis by the University of Waikato Radiocarbon Dating Laboratory, New Zealand) Discussion culture. In general, the deposits at Bulla support the A number of conclusions may be drawn from the observation of human habitation in particular ecological excavations at Bulla. First, the work confirms that there zones — on alluvial terraces in enclosed stream valleys are deeply buried archaeological deposits which lie sealed — following the Last Glacial Maximum (LGM) (Canning beneath sediments that have been put down during the 2009). More specifically, there is correspondence with past one to two centuries. Second, artefacts contained Keilor and Green Gully in terms of the soil contexts, within fine silt from below 1 m depth appear to be in situ, chronology of dated samples and artefact assemblages. which is suggested by consistency in the types of artefact The silt strata at Bulla, sealed below a metre of recent present and the soil conditions. Third, the materials and flood deposits, are comparable with soils of the Keilor morphology of artefacts from below 1 m, consisting Terrace: fine-grained silt which is alkaline and with predominantly of quartzite and featuring large cores, calcareous zones that contain biotic casts (Bowler 1970). blades and scrapers, suggest that the deposits may be of Formation of the Keilor Terrace began around 18,000 Pleistocene or early Holocene age. This chronology is BP, with the major phase of aggradation commencing confirmed by the age estimates generated from the two at approximately 15,000 BP and continuing up until the charcoal samples (12,142 cal. BP and 11,404 cal. BP). early Holocene, around 9,000 BP (Bowler 1970; Tunn Finally, artefacts from more than a metre below the depth 1998:37–38). The age estimate for the charcoal sample of the two charcoal samples are almost certainly older. It from the upper silt stratum at Bulla, which may represent is suspected that these deposits continue below the depth the younger ‘Intermediate Zone’ rather than the Keilor reached during the excavation. The yellow silt in Context Terrace proper, accords with this depositional history. 3 has a different character to the redder silt in Context 2, The age estimates from Bulla are slightly younger, where the two charcoal samples were collected, and most though broadly comparable with the age estimates likely results from a previous phase of deposition. (13,050–15,900 cal. BP) for charcoal samples from hearths Comparison of the results outlined above with situated within the Keilor Terrace Unit at Brimbank findings from elsewhere in the Maribyrnong River system Park (Canning et al. 2010:33; Tunn 2006:19–20). There reveals similarities in the excavated contexts and material are also parallels with the material culture present from

12 Pleistocene to Holocene in the Upper Maribyrnong River Basin: Excavations at Deep Creek, Bulla this equivalent context. Although variability in stone at a horizontal dimension, being unable to distinguish tool technology is a hallmark of Pleistocene and early the subtlety of shallow deposits, for testing deep vertical Holocene assemblages in Victoria (Bird and Frankel stratigraphy it can be effective. Since carrying out these 1998), and inconsistent identification and terminology excavations, the planned alignment of the bridge crossing make comparison of assemblages via published material of Deep Creek has been changed, and the area of Bulla 1 problematic, broad similarities with the Aboriginal will not be affected. places in Keilor can be discerned. These include a peak in artefact numbers at approximately 1–1.5 m below ground, predominantly manufactured on quartzite, in contrast Acknowledgments to the prevalence of finer-grained silcrete at shallower depths (Burke 1990; Canning et al. 2010:31; Munro 1998; We thank everyone involved with the investigations at Wright 1970). Likewise, artefacts from the lower strata Bulla: the project sponsor VicRoads and the Wurundjeri at Bulla are larger-sized and coarser-grained than the Tribe Land and Compensation Cultural Heritage Council. smaller and finer-grained lithic artefacts prevalent on the In particular we thank the Wurundjeri representatives surface and at shallow depths. who participated in the mechanical excavation: Jason The age estimates for the two charcoal samples from Tweedie (who operated the excavator), Wade Garvey and secure, artefact-bearing sediments confirm human Sean Wandin. occupation of the Deep Creek valley during the late Pleistocene to early Holocene. The apparently dramatic deposition of silt may be related to climate change References following the end of the LGM, which involved lower temperatures, decreased water flows and increased Anderson, W., V. Clark and A. Noble 2013 Bulla Bypass sedimentation (Bowler 1970; Canning 2009). A parallel and Melbourne Airport Link to Outer Metropolitan for this context may be drawn with the well-defined Ring: Historical Archaeology Survey. Unpublished Keilor Terrace Unit. The age estimates that confirm the report prepared by Dr Vincent Clark & Associates for antiquity of artefacts at a depth of 1 m were obtained VicRoads. from samples almost 2 m above the level of the deepest Bird, C.F.M. and D. Frankel 1998 Pleistocene and Early artefacts. Moreover, a lack of artefacts at depths between Holocene archaeology in Victoria. A view from 1.75–2.70 m suggests that the most deeply buried Gariwerd. The Artefact 21:48–62. artefacts (at 2.7–3.0 m) signify a different occupation Bowler, J.M. 1970 Alluvial terraces in the Maribyrnong horizon, of which only the uppermost level was reached. Valley near Keilor, Victoria. Memoirs of the National These results show that occupation of riverine terraces Museum of Victoria 30:15–58 in enclosed valleys during the late Pleistocene to early Burke, C. 1990 Analysis of the Lithic Assemblage from the Holocene was not confined to the main channel of the Keilor Archaeological Site. Victoria Archaeological Maribyrnong River, but is part of a wider pattern that Survey Occasional Report 30. Melbourne: encompasses the greater Maribyrnong River basin. Department of Conservation and Environment. There is considerable potential for investigation of these Canning, S. 2009 Broad scale palaeo-environmental contexts to shed light on cultural practices in relation to reconstruction of southern Victoria, Australia. The the landscape and environment during the Pleistocene to Artefact 32:4–14. Holocene transition. Canning, S., D. Griffin, V. Flynn and P. Ricardi 2010 Recent archaeological excavations of Pleistocene deposits at Brimbank Park, Keilor, Victoria. Postscript Archaeological Heritage 2(1):25–35. Crocker, S. and J. Wheeler 2013 Deep Creek – Weed While there can be little doubt that mechanical Removal, Revegetation and Rabbit Control, Bulla. excavation is a coarse-grained approach, the results CHMP number 12389. Unpublished report prepared of the investigation presented here demonstrate that by AHMS for Melbourne Water. reliable information can be obtained using this method. Duncan, J. 2001 Megafauna at Keilor and the timing of Careful operation of the mechanical excavator, controlled their extinction. Australian Archaeology 53:16–22. recording of soil contexts and artefacts, and the use of Frankel, D. 1982 Earth rings at Sunbury, Victoria. hand excavation to verify soil and artefactual conditions Archaeology in Oceania 17(2):89–97. at particular depths help to mitigate contamination Gill, E.D. 1966 Provenance and age of the Keilor Cranium: between stratified layers. Under the circumstances and oldest known human skeletal remains in Australia. constraints of the project, mechanical excavation was a Current Anthropology 7(5):581–584. suitable tool to address specific research questions, and Gregory, J.W. 1903 The Geography of Victoria: Historical, although it is poorly suited to general subsurface testing Physical and Political. Melbourne: Whitcombe and

13 William Anderson and Racheal Minos

Tombs. Wildwood Road Soil Removal Project near Deep James, A.V.G. 1920 The physiography and geology of the Creek, Bulla. Unpublished report prepared for David Bulla-Sydenham area. Proceedings of the Royal Society J. Allan and Associates. of Victoria 32(2):323–349. Rossiter, A.G. 2003 Granitic rocks of the Lachlan Fold Moloney, D. 1998 City of Hume Heritage Study: Belt in Victoria: potential keys to the composition of Former Shire of Bulla District, 1998. Volume Two: the lower-middle crust. In W.B. Birch (ed.), Geology Environmental History. Unpublished report prepared of Victoria, pp.217–237. Sydney: Geological Society for Hume City Council. of Australia. Mulvaney, D.J. 1970 Green Gully revisited: the later Sutherland, P. and T. Richards 1994 Shire of Bulla excavations. Memoirs of the National Museum of Aboriginal Archaeological Study. Sunbury: Shire of Victoria 30:59–77. Bulla. Mulvaney, D.J. 1998 Dr Gallus and Australian Tulloch, J. 2003 An Archaeological Survey at Lancefield archaeology. The Artefact 21:4–8. Road, Sunbury East, Victoria. Unpublished report Munro 1998 The stone artefact assemblage from Keilor. prepared by Biosis Research for Austcorp Group. The Artefact 21:19–34. Tunn, J. 1998 Pleistocene landscapes of Brimbank Park, Presland, G. 1983 An Archaeological Survey of Keilor, Victoria. The Artefact 21:35–47. the Melbourne Metropolitan Area. Victorian Tunn, J. 2006 An Aboriginal campsite on the Maribyrnong Archaeological Survey Occasional Report Series 15. River – new dates for Keilor. The Artefact 29:14–21. Melbourne: Ministry for Planning and Environment, Vines, G. 1995 Grey Box Forest Ecological and Cultural Victoria. Heritage Project, Cultural Heritage Study and Rhodes, D. 2000 An Archaeological Survey of the Management Proposal. Unpublished report prepared ‘Riverlea’ Property, Konagaderra Road. Unpublished by Melbourne’s Living Museum of the West for the report prepared by Terra Culture for Ecology Federal Airports Corporation. Australia and Tract Consultants. Wright, R.V.S. 1970 Flaked stone material from GGW-1. Richards, T. and P. Sutherland 1995 Surface and Sub- Memoirs of the National Museum of Victoria 30:79– Surface Archaeological Survey of the Proposed 92.