Bagot's Mill: Genesis and revelation in an archaeological research project GRAHAM CONNAH Emeritus Professor, University of New England Visiting Fellow, Australian National University The Bagot's Mill research project set out to investigate the ruin of a watermill at Ben Lomond in northern New England. Work commenced in 1979 but the fieldwork, excavations, archival searches, and artefactual analysis, continuedfor some years. The original aim ofthe project was to test an hypothesis generatedfrom the documentary sources: namely that the builder ofthe mill, Christopher Thomas Bagot, had made a mistake in building it at this place and that the mill would not have been viable in the long term even ifhe had avoided bankruptcy soon after its construction. Structural and stratigraphic evidence on balance supported that hypothesis but analysis of the excavated movable artefacts revealed that most of them dated from after the abandonment ofthe mill. This archaeological evidence led to the generation ofa second hypothesis: that such an abandoned site is likely to continue to attract human use, although the character ofthat use may change. The results ofthe artefactual analysis, which suggested that the ruined structure had been usedfor garbage disposal, for target-shooting and as a place for children to play, clearly supported that hypothesis. It showed that the whole ofa site's life history must be taken into account when trying to understand it. Implicit in the outcome of the Bagot's Mill project, however, is the need for flexibility in research design: the unexpected might be inconvenient but it may have great significance. INTRODUCTION LOCATION, ORGANIZATION AND CONDUCT OF THE RESEARCH PROJECT This is the story of a research project in the field of Australian historical archaeology. As every story should, it has a moral. It The project in question concerned the site and vestigial ruin of concerns an investigation conducted over a period of many a watermil1, situated on Ben Lomond Station, a little north of years, whose ful1 publication has been long delayed. The Guyra and about halfway between Armidale and Glen Innes, in excavations which provided the core of the relevant data were the New England district of northern New South Wales carried out in 1979, 1980, and 1982 respectively. The analysis (Fig. 1). The site lies in the upper val1ey of the Mann River, on of the excavated evidence was done mainly during the years Portion 165, in the Parish of Bagot, County of Clarke (NSW 1984 to 1989. Publication was postponed because of the 1962), at map reference 78627612 (NSW 1975a). This is part writer's involvement with several other research projects, of the sparsely populated highest country on the New England particularly from 1987 til1 early 1996. Such a long gestation Tablelands, the miII site being at just below 1340 m above sea has resulted in an outcome different from what might have level. The immediate area is one of Tertiary basalts on which been had the research been concluded mqre expeditiously. are relatively fertile soils, although loose surface stones and Evidence which at the time of the excavations seemed of little outcrops occur on higher ground. A substantial number of trees or no relevance to testing the hypothesis which initiated the (mostly stringybark) survive but much of the land has been investigations, was found to be of value for testing a new and cleared for a century or more and provides classic open upland different hypothesis. The project had commenced by asking sheep grazing, although large numbers of cattle are also historical questions of archaeological data, and indeed it did present (Lea, Pigram & Greenwood 1977). prove possible to answer those questions to some extent, but it The writer first visited this site on 21 February 1977, it concluded by also asking archaeological questions of the being drawn to his notice by Graham Walker, at that time of archaeological data; a shift which to an archaeologist is of the Department of Geography at the University of New considerable interest. After al1, if the only thing historical England. Subsequently Jil1ian Oppenheimer, then of the archaeologists can do is write history, then they may as wel1 be Department of History at the same university, conducted historians. lengthy archival searches in Australia, Britain and Ireland for The real issue, however, is whether such a shift should information about the site and the people associated with have been al10wed to occur. Three decades of archaeological it. Survey and excavations were then carried out by the writer concern with the hypothetico-deductive approach have surely for two weeks in August-September 1979, two weeks in May demonstrated the need for a greater level of rigour than this? 1980, and three weeks in August-September 1982. He was Perhaps, but perhaps also a little flexibility is crucial in the assisted by Douglas Hobbs in 1979 and 1980 and on each conduct of any research, so that one is able to restate, augment of the three occasions the excavation workforce was made or even replace an hypothesis or hypotheses as the up of volunteer undergraduate students from the Department investigation proceeds. Non-archaeological critics may of Prehistory and Archaeology (now the Department of sometimes feel that archaeologists rarely find what they are Archaeology and Palaeoanthropology) at the University of actual1y looking for but a more real concern is whether New England. The size of the excavation team varied from ten archaeologists are likely to find what they are not looking for. to twenty people on average and work was conducted Rigour should not require the wearing of blinkers, at times the according to a tightly scheduled programme. Al1 the field seemingly irrelevant may prove ofgreater significance than the photography (including aerial work) was done by the writer, relevant. Rigour may be desirable, rigor mortis is not. who also attended to the detailed excavation recording; Douglas Hobbs did the surveying and planning. After the excavations were over, a field survey of the mil1 race, and of 3 QUEENSLAND '.-._ ....... - ........ N t Tasman Sea D 0 - 300m 300 - 1000m III above 1000m o 50 100m DH 96 Fig. ]: Location ofBagot's Mill. the dam on Llangothlin Lake, was undertaken by Kathy Panitz The Bagot's Mill project was carried out while the writer in 1989, and the analysis of the excavated evidence by Sue was Associate Professor, and subsequently Professor, of Pearson with some work also done by Stuart Murray. In short, Prehistory and Archaeology, and then Archaeology and this research project, like many another in Australian Palaeoanthropology, at the University of New England. It was archaeology, was a team effort, with most participants purely a research project with no elements of heritage volunteering their services. It was also greatly assisted by the consultancy. Its costs were modest, mainly involving generosity of local people, particularly John Creagan of transportation, subsidized food for excavation volunteers, Moredun Station and Dick Streeter of Ben Lomond Station photographic processing, and some paid analysis work. who provided accommodation for the research team free of Funding came from the University of New England Internal charge, the former in 1979 and 1980, the latter in 1982. In Research Grants, with a small input from an Australian addition, mention must be made of Bill Cameron of the Glen Research Grants Committee grant that was awarded to the Innes and District Historical Society, who gave assistance with writer and Alan Jones, of the Department of Geography at the local archival material and generously shared his extensive University of New England, for a llirge-scale archaeological knowledge of the area. Arnold Goode of the Uralla Historical aerial photography project, of which the mill formed only a Society also helped significantly with archival sources, by tiny part (Connah & Jones 1983). Considerable infrastructural making available substantial unpublished information that he support was also provided at all stages of the project by the had collected concerning H. S. Goddard, the engineer who University of New England, particularly by the writer's own designed and constructed the waterwheel. department where Douglas Hobbs and Malcolm Abel were Archival work continued, indeed, for some years after the especially helpful, and by the Media Resources Unit whose completion of the excavations, and the writer is much indebted photographic staff did artefact photography and processed to the Archives Office of New South Wales, in Sydney, and to numerous films and prints over the years. the Westpac Banking Corporation Archives, also in Sydney, for The foregoing details have been included here, rather than access to information concerning Bagot's insolvency, which being placed in the customary manner in an appended was such a crucial event in the history of the mill. Also long 'Acknowledgements', because they show how this project was after the excavations, Warwick Pearson was able to set Bagot's organized and conducted. As in any research, this provides an Mill in context with his study of a selection of waterrnills in essential basis for assessing what follows. both Australia and Britain from the point of view of technology transfer (Pearson 1995; in press). THE GENESIS OF A QUESTION: dairy, where I saw on the shelves some hundreds of the BAGOT'S MISTAKE? famous Ben Lomond cheese. In them a large trade is done yearly all around New England. A better or a The central figure in our story has to be Christopher Thomas more marketable article could not be wished for. Bagot, born in 1826 presumably at Fontstown, County Kildare, Ireland, where his father was an Anglican clergyman, died on People around do say that the Ben Lomond herd is 27 January 1894 at Broadwater, on the Richmond River, a good one. I saw some very fine pure-bred Herefords northern New South Wales (Walker 1963; Oppenheimer and Devons-a bull and three heifers of each lately 1980-1981; Principal Registrar NSW 1979).