Last updated: 14th June 2008 (R.Tabor)
The excavations of 1m squares down to the geological surface at every 100m within sampled areas was at once the most revolutionary and the most generally informative technique introduced by the project, providing a relatively small but very high quality prehistoric database. The link between test pits and geophysical data was established more firmly by the introduction of pits targetting key geophysical anomalies at the expense of ploughzone sampling. Evaluation and theoretical discussion of this strategy appears in The South Cadbury Environs Project, Neolithic to Medieval landscapes. Volume 1: Methods, themes and synthesis and in Regional Perspectives in Archaeology: from sampling to narrative.
The programme of regular test pitting was introduced in a pilot study on the Sigwells
plateau in February 1998. It was so successful that it became, with geophysical survey,
the core of the project. Originally intended as a test of the relationship between
surface and the subsurface finds, it became the principal means for dating the distribution
of activity in the prehistoric and Romano-
The project began with very ad hoc gridded fieldwalking, targetting areas thought promising. Eventually it moved on to corridor walking, a system using 1m wide contiguous 20m long strips. This is now restricted to surface collection on some plateau areas.
After its first successful use during the Sigwells pilot study shovel pitting (sieving a fixed amount of soil at regular intervals) was employed to reduce variables associated with the surface conditions and performance area. Ultimately, like fieldwalking, it proved ineffective on middle to lower valley sides and bottoms and open plain due to soil movement.
This shallow test pit on a low plateau illustrates extreme plough damage. The surface of the field was littered with stone where natural rock had been shattered and churned up by the soil. The soil have been 40cm or more deeper during the Neolithic.
This was one of the deepest regular test pits which reached natural (in some instances the geological surface was below a depth at to which it was considered safe to dig). The gravelly layers are partly for Medieval and Roman track metalling,, partly from redeposited natural. Below that are Bronze Age hillwashes sealing an early soil.
When surface collection methods were rejected targeted test pits replaced them. They introduced a substantial bias into the sampling but made it possible to date extensive boundary systems detected by geophysical survey.
