Broom Lower Palaeolithic Site (Part 2 of 2)

The Broom Sediments

The deposits at Broom form part of the Axe Valley Formation, and indicate at least three periods of aggradation. The well-bedded Holditch Lane Gravel Member is the lowest and earliest deposit. The character of these sediments suggests deposition in cool or cold climatic conditions. There is no unequivocal evidence that any of the artefacts found at Broom came from the Holditch Lane Gravel.

Overlying the Holditch Lane Gravel is the Wadbrook Member of the Axe Valley Formation (Figure 3). broom3In the Railway Pit and Pratt's New Pit the Wadbrook Member consists mainly of horizontally bedded silts and clays but in Pratt's Old Pit it was largely gravel. These sediments are interpreted as, respectively, the floodplain and channel deposits of the rivers Axe and Blackwater. The pollen recovered from the fine-grained sediments in the Railway Pit is indicative of a post-temperate flora, with boreal elements combined with more temperate taxa. A late MIS 9 age is supported by the OSL dating of the sediment itself. The Wadbrook Member has been the source of most of the Palaeolithic artefacts found at Broom. The distribution and condition of the artefacts suggests that they were discarded on the floodplains of the Axe and the Blackwater near the confluence of these rivers, or on land-surfaces close to the rivers.


Figure 3: Exposure of the Wadbrook Member (Silts & Sands Bed), Axe Valley Formation (image copyright, Dr Chris Green)

broom4Overlying the Wadbrook Member is the Fortfield Farm Gravel Member (Figure 4). This is more variable and less evenly bedded than the Holditch Lane Gravel, although its general character also suggests deposition under cool or cold climatic conditions. Palaeolithic artefacts were undoubtedly recovered from the Fortfield Farm Gravel, but in small numbers. OSL dating of sand bodies within the Fortfield Farm Gravel suggest that the date for its deposition was probably within MIS 8.


Figure 4: Exposure of the Fortfield Farm Gravel Member, Axe Valley Formation (image copyright, Dr Rob Hosfield)

The Broom Artefacts

Analysis of the principal collections (at the Dorset County Museum, Dorchester; the Royal Albert Memorial Museum and Art Gallery, Exeter; and the British Museum) and additional, smaller collections at other local and regional museums indicates the existence of at least 2,300 artefacts from Broom, although the complete total was certainly larger. The overall assemblage is dominated by bifaces, with only small numbers of flakes, cores, retouched flake tools, and other items, including very occasional Levallois pieces. Comparison of the Broom bifaces with other British Lower Palaeolithic assemblages suggests a refined industry, despite the artefacts' frequent lack of symmetry. The mainly used Greensand cherts were clearly not a barrier to well-made bifaces. The majority of the bifaces are ovate or cordate in form, although the assemblage is also notable for the variety of forms produced. If the OSL chronology (see above) is reliable, then Broom is biface-dominated at a time (MIS 9/8) when Levallois technologies were being introduced in south-eastern England. Roughly one quarter of the Broom bifaces are distinctive for their characteristic asymmetry (Figure 5). Analysis of raw materials, metrics, typology, and technology suggests that this asymmetry was not solely, and possibly not primarily, a by-product of prosaic factors such as blank size/shape or re-sharpening intensity. Instead we propose that it was a deliberate product of hominins' idiosyncratic choices and/or traditions, operating both amongst individuals and at the scale of the group. This is supported by the apparent association, primarily documented through Bean's records, of the majority of the artefacts with the Wadbrook Member, probably of late MIS 9 age - which we have interpreted as the result of a single, continuous, perhaps multi-generational, phase of Lower Palaeolithic occupation at Broom.


Figure 5: Broom biface with the characteristic 'lop-sided' asymmetry (image copyright, Dr Rob Hosfield)

The physical condition of the artefacts suggests that the majority are locally derived, although probably not in situ. The attractions of Broom and the Axe and Blackwater valleys are likely to have included fresh water, lithic raw materials (from the river gravels, valley sides, and interfluve plateaux), game, and perhaps also a crossing point over the rivers. Interestingly the size of the Broom artefact assemblage is currently unique in the south-west region. This pattern may reflect the distribution of favourable raw materials, but the distinctiveness of Broom both within and beyond the Axe Valley may also reflect local and regional taphonomic factors related to terrace formation and preservation.

Broom and The Axe Valley

It is apparent that the River Axe had already cut down well below the level of its modern floodplain in MIS 10 and that sediments of MIS 9 age, the Wadbrook Member of the Axe Valley Formation, overlap the height range of the modern alluvium. These findings give the Axe valley an identity that is almost certainly unique among rivers of comparable size draining to the south coast of England. We propose that the highly erodible bedrock of the valley floor, combined with the ready availability of an erosive bedload, is the most likely cause of an unusually rapid lowering of the valley floor during Middle Pleistocene episodes of low sea level, a pattern almost certainly intensified by the catastrophic enlargement of the Channel River. As a result of this accelerated downcutting the terrace long profile in the Axe valley, particularly in its lower and middle reaches, is systematically lower than the profiles of terraces of the same age in neighbouring valleys. An associated consequence of the erodibility of the bedrock is the susceptibility of terrace remnants to erosion and hence the unusually patchy preservation of terraces in the Axe valley.

C.E. Bean

Our understanding of Broom owes a considerable debt to the fieldwork and archive of Charles Bean. Bean undertook a large number of visits to Pratt's Old Pit throughout the 1930s, compiling a mixture of more and less detailed field records of the pit's sediments, stratigraphy, development, and artefact provenances. Alongside the written record Bean also produced an invaluable range of illustrations in the form of photographs, sketches, and plans. Moreover Bean's professional duties as a surveyor allow a degree of confidence to be placed in his field record. Without the presence of Bean at Broom it would seem likely that many more of the artefacts collected by the quarrymen would have been widely dispersed into a number of different collections, or perhaps ignored entirely, making any study of the site or the artefacts found there much more difficult, if not impossible.

There are, nonetheless, difficulties associated with the interpretation of the Bean archive: in particular, relatively few of the recovered artefacts from Pratt's Old Pit have a detailed provenance. In addition, even for a researcher familiar with them, Bean's notebook entries can sometimes be ambiguous, while it is a matter of regret that many of the flakes and cores recorded by Bean became separated from the bifaces and no longer form part of the collection.

Despite these limitations, there is one simple way of measuring Bean's contribution to our understanding of the Broom site - by comparing what is currently known of the sediments, stratigraphy and archaeology of Pratt's Old Pit, despite the fact there are now almost no exposed deposits, with what is known of Pratt's New Pit and the Railway Pit, where more substantial exposures remain: there is very little contest, and current understanding of the Broom geology and archaeology, with especial reference to Pratt's Old Pit, would be considerably the poorer without Bean's notebooks, photographs and artefact collection to underpin the more recent investigations.


The three main conclusions from our research are that:

  • The majority of the Lower Palaeolithic artefacts preserved at Broom represent evidence of occupation during a single, relatively short-lived, phase of favourable environmental conditions.
  • The distinctive features of the Broom Lower Palaeolithic (Acheulean) assemblage owe more to local group traditions than to any environmental constraints or wider/longer-term technological patterns.
  • The distinctive character of the Quaternary stratigraphic record in the Axe valley reflects, firstly, an intensification of geomorphological processes in the Middle Pleistocene following the enlargement of the Channel River; and secondly, a local bedrock lithology that favoured a rapid and far-reaching fluvial response.

For full details of the Broom research, please see our recent monograph (from which the above summary text is drawn):

Hosfield, R. & Green, C. P. (eds.) 2013. Quaternary History and Palaeolithic Archaeology in the Axe Valley at Broom, South West England. Oxbow: Oxford.

Rob Hosfield & Chris Green, April 2014.




Things to do now

Read the introduction
Read about the investigator

Follow us

Contact us

Page navigation


Search Form

A-Z lists