The Palaeolithic collections of the Wiltshire Museum are relatively limited. There are 1,248 entries attributed to this period in the collections management database, with the majority of these being records of individual Lower Palaeolithic handaxes. Whilst there have been some recent acquisitions of chance finds, such as a handaxe from Huish reported through the Portable Antiquities Scheme (DZSWS:2019.10), the majority derive from old collections. In both cases there is limited surviving contextual information.
By far the most significant assemblage of Palaeolithic objects derive from the artefact-rich gravel pit at Knowle Farm, Little Bedwyn. 1,132 of the records are attributed to this site, and it is likely that some of the handaxes attributed to neighbouring parishes, such as two from Savernake, may also have derived from the site or a related deposit. The Knowle Farm gravel pit is famous for the quantity of flint recovered, and by 1903 over 2,000 flint ‘implements’ had reportedly been discovered (Cunnington and Cunnington 1903). This represents the most significant deposit of Lower Palaeolithic material in the region, and whilst the handaxes are now widely dispersed, the collection held in Wiltshire Museum remains the largest (Roe 1968; 1969). The collection has been recognised from early on as a mixture of multiple deposits, probably deposited by river action, unfortunately limiting their usefulness for statistical analysis (Cunnington and Cunnington 1903; Roe 1968; 1969).
Several descriptions of the site appeared in the Wiltshire Archaeology and Natural History Magazine (WANHM) in the early 20th century, however they contain insufficient detail to allow for in depth discussion of the geology or archaeology of the site (Cunnington and Cunnington 1903; Dixon 1903; Kendall 1906). More recently, in 1977 a trial trench was opened by mechanical excavator (Froom 1983). This was able to provide limited clarification of the clarification of the site, but due to the method of excavation the stratigraphic relationships of the 70 Paleolithic flints recovered were not recorded, with the exception of a single handaxe (Froom 1983). It is also disappointing that none of the material recovered during this excavation appears to have entered the museum collections. Both Kendall (1906) and Froom (1983) note the presence of flakes and other evidence of knapping within the material recovered, although this is denied by the Cunningtons (Cunnington and Cunnington 1903). A sample of 461 of the less worn handaxes were examined in detail by Roe for his PhD, who noted that the assemblage is dominated by ovate forms, and suggested that it was characterised by unusually crudely-made tools (1968; 1969). The Knowle Farm flints are also known for a highly distinctive and poorly understood ‘gloss’ (Cunnington & Cunnington 1903; Dixon 1903).
The only other sites associated with significant numbers of Palaeolithic objects are both in the Salisbury-area, with 58 handaxes attributed to the gravel extraction pits at Millford Hill and Bemerton. The handaxes were donated by C.J. Read, who also published the sites (Read 1884), and whilst detailed sketch plans allow the deposits to be placed on the map with relative accuracy, stratigraphic detail is again limited. Examination by Roe (1968; 1969) also suggests that the groups are unlikely to represent a closed group of implements, reducing their usefulness for statistical analysis.
Comparison of the distribution of the findspots of Palaeolithic objects in the museum collections with those plotted by Roe (1969) reveals little change in the second half of the 20th century. The Kennet Valley and Marlborough- area in North Wiltshire continues to form the focus of the distributions, with a developing scatter of chance finds and stray flints in the North West of the county.
The Mesolithic collections of the museum are similarly limited, as is the case for the archaeological record for the period in the county as a whole (Hosfield et al. in Webster 2007). There has been only a single significant excavation of an in situ Mesolithic site since Radley’s (1969) review of the period, at Blick’s Mead, Amesbury, south of the Museum’s collecting area (Jacques and Phillips 2014). Searching the collections management database produces 1,908 records, however, this number is not very informative. Most again record small groups or single flint flakes or tools, often as part of larger, mixed field walking assemblages. A review of the collections reveals just 13 groups of more than 50 flints identified as belonging to the Mesolithic (Figure 2.1). Twelve of these groups were collected through fieldwalking or survey, and there is a notable number of assemblages of 200-800 flints in the north west of the county, as well as a collection of 776 flints collected during a survey by Gingell in Teffont (Gingell and Harding 1983). A smaller assemblage of 282 Mesolithic flints is attributed to Golden Ball Hill, Alton, where a programme of geophysical survey and trial excavation by Cardiff University in 1997 identified in situ Mesolithic occupation levels (Dennis and Hamilton 1997). Similarly, a small field walked assemblage of Mesolithic flint is attributed to Hackpen Hill, also a known Mesolithic site (Whittle 1990: fig. 2).
The most significant assemblage derives from the excavations at Oliver’s Hill Field, Cherhill, where excavations in advance of development in 1967 identified occupation spanning the Late Mesolithic to Early Bronze Age (Evans and Smith 1983). Although thin and patchy, and in places cut by later ditches, the site was well stratified with Mesolithic layers sealed by a deposit of tufa. The latter contained a lens of charcoal near its base radiocarbon dated to 5280 +/-140 BC, as well as smaller quantities of Mesolithic flint and bone. No precise count of the Mesolithic flint assemblage was published, however, it was estimated to comprise c. 10% of the 130kg of struck flint recovered from the site. It is dominated by bladelets and contains both scalene micro-triangles and obliquely blunted points, and was argued to represent a single broadly contemporary industry, with most flints described as being in fresh condition. In addition to the flint assemblage, a potentially important assemblage of 1,681 animal bone fragments were recovered from Mesolithic contexts. Of this group, only 125 fragments were positively identified due to the extent of fragmentation, and the assemblage was not quantified beyond NISP, it was however not deposited with the rest of the archive at the Wiltshire Museum, and the osseous material was deposited with the British Museum (Natural History) under the accession numbers ARC 1981.5163-5533 and ARC 1982.5003-5016. Overall, only a single Mesolithic feature was identified, a ‘working hollow’ (although see Davis 2012, cited below), and the original excavators interpreted a general trend of gradual abandonment as the site became increasingly saturated with water.