Roads were the economic arteries of Roman Britain and afforded obvious commercial advantages to the settlements along their course. Such settlements occur in a great variety of forms and sizes, although there has perhaps been an over-emphasis on morphology rather than on understanding the activities performed in these places. This slim monograph presents a highly professional account of the results of a very spatially limited investigation in advance of cable-laying across the line of the road from Bath to Silchester. The prime discovery was a previously unknown roadside settlement, although it is difficult to say much about its form given the narrow strip examined, which, to judge from the distribution of metal detecting finds, lay near the edge of the settlement. Finds extended for around 900 m along the road, and the overall area of activity, if not actual settlement, is estimated to be in the order of 12–20 ha. One of the fields in the heart of the finds scatter has the suggestive field name ‘Blacklands’ (frequently an indicator of the site of ancient habitation). Two buildings were recognised, but neither was a classic strip building fronting on to the road. One was of semi-cellared form and joins the growing number of these structures now recognised in various parts of southern Britain from the late Iron Age onwards. The large number of ovens, 41 in total, within the excavation area is noteworthy.

It is inevitably difficult to contextualise the structural evidence from restricted investigations such as this, and the particular interest of this report derives from the study of the artefacts and biological evidence. Pigs are unusually well represented in the faunal assemblage, with many of the bones burnt and recovered from pits adjacent to the ovens. It appears that young succulent pigs were roasted in the ovens, and, once the meat had been stripped from the carcass, the waste debris was left to incinerate in the oven before being raked out and dumped in the pits. The pork products were likely exported along the road, perhaps to Bath, although they could also have supplied a roadside delicatessen.

Beanacre lay close to the point where the road crossed the Avon, and the river crossing was surely influential in site selection. The settlement lay only 3.5 miles along the road from Sandy Lane and 9 miles from Bath. The proximity to Sandy Lane is curious. Occupation at Beanacre commenced in the late first or early second century a.d. and continued into the second half of the fourth century. Occupation was therefore contemporary with Sandy Lane rather than sequential. Sandy Lane is conventionally identified as Verlucio of the Antonine Itinerary, which is recorded in Iter XIV of the Itinerary as lying 15 Roman miles from Bath and 20 from Cunetio (Mildenhall). The overall distance between these two places given in the Itinerary is thus 35 miles, whereas the true mileage is 32, so some inaccuracy is represented. A.L.F. Rivet and C. Smith (Place Names of Roman Britain, London 1979) suggest that the 20 miles from Verlucio to Cunetio should be amended from xx to xv to produce a more reasonable figure for the stage. The identity of Sandy Lane as Verlucio is not beyond question therefore, especially as we know so little about that place (a defended enclosure of c. 3 ha is apparently represented, so Sandy Lane could well be a smaller settlement than Beanacre). Beanacre demonstrates the potential for the discovery of previously unrecognised roadside settlements and cautions us against economic models that assume we have a complete knowledge of all such settlements. We clearly don't.