This third and final volume of the excavations completes the record of 49 years of examination of prehistoric, Roman and post-Roman settlement in the parish of Frocester in the Gloucestershire Severn Valley. Eddie Price has rightly been lauded for his dedication to the archaeology of his farm, and his insistence on high standards in fieldwork and reporting. This volume maintains the quality of the previous two reports and is a credit to both P. and his helpers from the Gloucester and District Archaeological Research Group. It reports on a series of trenches dug around the periphery of the Frocester Court farmstead and villa which were located to check the results of a magnetometer survey. Most of the work took place in front of the late third- and fourth-century villa and revealed small enclosures, roundhouses, ovens and furnaces. A small third-century industrial complex housed in a timber building continued in use well into the fourth century and thus spans the period before and after the construction of the villa house. It may have been replaced by a new complex which P. reasonably interprets as a brewhouse involved in the domestic production of ale.

Frocester Court is in many ways an unremarkable villa, and it is perhaps most widely known for its post-Roman evidence about which a little more can now be said. In Volume 1 of the excavations mention is made of a pagan Saxon bead and possible claw-beaker found near to the villa house. These pieces are now published in full (nos 25 and 63) and in both cases we find that they need not necessarily be later than the late fourth century. P. has always been cautious in his interpretation of the post-Roman evidence, which could date some good while after the villa house burnt down. While developer work over the last 20 years in the Cotswold-Severn region has routinely turned up grass-tempered pottery in the uppermost deposits on Roman rural sites, in every case the nature of the occupation associated with this pottery is extremely unclear. The collection of post-Roman buildings at Frocester Court therefore remains our best evidence for structures in the region outside of the Anglo-Saxon architectural tradition, although we must hope that the post-Roman occupation at Crickley Hill will one day be finally published in the detail it deserves.

A notable discovery reported in this volume is an inhumation burial with hobnails around its feet which produced a radiocarbon date of 715–1016 cal a.d. at 95.4% confidence. Burials accompanied by hobnailed footwear from places as far apart as the small town at Shepton Mallet, Somerset, the early Christian cemetery at Llandough, Glamorgan and the probable villa at Parlington Hollins, West Yorkshire, have also all now produced fully post-Roman radiocarbon dates, so we must be wary of assuming that such burials, without other dating, are necessarily late Roman. Interpretation of this burial is unclear. On the one hand it raises the possibility that some of the post-Roman occupation at Frocester Court could be as late as the eighth or ninth century, but on the other the burial could suggest that occupation had ceased by this date and that it is part of a well-attested tradition of post-Roman burials dug into the ruins of long-abandoned villa houses. If this were so, however, the lack of burials in the ruins of the main villa house is slightly surprising. Much fresh information could usefully be gained from a future programme of linked radiocarbon dating of animal and human bones from the post-Roman sequence.

We might reasonably expect after all the investigation that has taken place that we would now have a reasonable understanding of the economic base of the settlements. P. has previously concluded that the economies of the farmstead and later villa were both founded on mixed farming. In the third century the farmstead may have been involved in the collection, storage and distribution of agricultural produce, but as none of the fourth-century structures were directly associated with farming it is possible that by this time those activities had moved to outlying locations away from the villa house. Once the house was constructed more diverse activities took place in its environs, such as metal working, antler working and brewing. While the villa house might have been built on the profits of farming, once established within its ditched enclosure (parkland?) the main business was now seemingly carried out elsewhere. P.'s intimate knowledge of the land he farms makes these observations on how the settlements sat and interacted with their landscape both fascinating and utterly convincing. This volume brings the 49 years of investigation to a highly satisfactory conclusion. P. richly deserves his recent award of a MBE.