- Based on an article by Peter Beaumont1 and John McNabb2
Canteen Kopje has been mentioned in the archaeological literature for almost a century. Eminent prehistorians including C. van Riet Lowe, the Abbe Breuil and J. Desmond Clark visited and described it, yet only recently has the site been investigated systematically.
The recent research was made possible by a grant from the Royal Netherlands Embassy which (together with support from the US Embassy) also provided for development of information displays at Canteen Kopje and the creation of a new Barkly West Museum. None of this - research or museum development - would have been possible without the support of the community. When in 1997 pressure was exerted for Canteen Kopje - a national monument - to be deproclaimed for renewed diamond digging, the community recognised the value of conserving heritage for education and tourism.
At issue was more than just the principle of preserving a declared site. It is true that by the 1960s the significance of Canteen Kopje had seemed somewhat diminished. The validity of the elaborate Vaal River sequence of alluvial deposits, and their correlation with past climates and a succession of archaeological "type fossils", had been called into doubt. Besides, accumulations of stone tool assemblages in alluvial settings (i.e. deposited by a river, thus not in primary context) were of limited archaeological meaning. But the first systematic archaeological investigation of Canteen Kopje itself (by Beaumont, from the late 1980s), drawing also on the detailed geological assessments by Helgren and by de Wit, has led to a revision of that pessimistic view. The site is now known to be characterised by a stratified deposit with distinct gravel or, rather, rubble units - containing artefacts - that are mainly of colluvial (as opposed to alluvial) origin, and possibly spanning Lower to Middle Pleistocene times. These units are capped by wind-blown Hutton Sands. An estimate based on a sample since recovered from the site indicates that it must contain somewhere in the region of 100 million stone artefacts. The site is archaeologically significant, and the bid to save it from being mined away entirely was far from trivial.
Excavations were located in two parts of the site: one within the declared area (Area 1), and one outside of it (Area 2, since completely mined away). The excavated samples from Stratum 1, Stratum 2a and Stratum 2b Upper have been analysed by John McNabb, while those from Stratum 2b Lower have been examined by Peter Beaumont, but remain to be classified. From the sections in the excavations and in several old diggers' pits, the overall stratigraphy of the site has been deduced:
Stratum 1: Wind-blown Hutton Sands that extended over much if not the entire site. Horizontal palaeosols and lenses within this unit - currently awaiting further investigation - represent four successive land surfaces, which signify resting phases in the deposition of the unit. Later Stone Age artefacts occur on the surface, while MSA and Fauresmith are present within the Hutton Sands and at the interface with Stratum 2a.
Stratum 2a: This unit (1.8 m thick) appears to have covered all of the site beneath Stratum 1, and is the unit traditionally associated with Canteen Kopje. It was formerly characterised as Younger Gravels II and as Rietputs A. This rubble is made up of clasts (broken-down rock), mainly of cobble size, in a matrix of pale orange sand.
Stratum 2b: This unit, found in the excavations in Area 1, was previously unrecognised, and consists of Upper (1.4 m thick) and Lower (5.6 m thick) sub-units, with stratification within them. Stratum 2b rests on andesite bedrock.
The "gravels" at Canteen Kopje are interpreted as a colluvial lobe built up largely by gravitation from the adjacent hill (as previously posited by Alex du Toit and Helgren). There is little marked abrasion on andesite cobbles and artefacts within it. In section, the rubble surface is not flat, but dips perceptibly from east to west, with evidence of two possible pulses of accumulation, in which Stratum 2b represents the first, and Stratum 2a, a second more extensive accretion over the entire site. Small, well rounded, non-andesite pebbles within these units are derived from vestiges of the Older Gravels on the crest of the adjacent hill. It is suggested that the formation of this deposit took place largely during arid periods, when hillside sediment and disintegrating bedrock was only loosely held by reduced vegetation cover - and then gravitated downwards as a result of sporadic torrential downpours.Analysing the archaeological material from Canteen Kopje (Beaumont & McNabb 2000), John McNabb notes that "the vast majority of artefacts, in any raw material, were only slightly worn. This is particularly true of the andesite pieces [which predominate]. These can not have travelled very far from their point of original manufacture...None of the artefacts showed any sign of the edge chipping and pitting that is characteristic of transport by water."