Why should readers of a journal concerned with Roman Britain be interested in the largest prehistoric mound in Europe? Silbury Hill was constructed roughly two and a half millennia before the Conquest, and until comparatively recently its main significance for specialists on the later period was its relationship to the Roman road between London and Bath. In fact it was the demonstration that this road skirted the foot of the mound that established the prehistoric credentials of Silbury Hill. It took an excavation in the nineteenth century to show that it had not been buried beneath the earthwork.

The title of the book justifies its inclusion among the reviews in this journal. This is the story of Silbury Hill: a narrative that is not confined to the prehistoric period. The first chapter reinforces the message for it is called ‘The nature of time’. The book traces the use of this remarkable place from its first discovery by antiquarians to the beliefs of alternative archaeologists in the present day. Do not be put off by this. This is not a work of abstract theory. It is lavishly illustrated, thoroughly accessible and very clearly written.

If it traces the structural history of the mound and its surroundings, it also traces the history of fieldwork and interpretation. It encapsulates many different strands in the development of archaeology as a discipline, from John Aubrey to Flinders Petrie, and from the excavation sponsored in the 1960s by the BBC to the distinguished fieldwork conducted by the authors nearly 40 years later. That is appropriate, for each generation has left its own legacy. The poorly reinstated excavations of the nineteenth and twentieth centuries threatened the stability of the mound, and it was this problem that led to new work at the site. The inadequate records of the excavation conducted by Atkinson necessitated a new and more detailed assessment of the deposits that still survive. Not surprisingly, it has led to the rejection of his original interpretation, which was only partly published in his lifetime.

The Roman period at Silbury has undergone a similar reassessment. Artefacts and structures had been identified around the foot of the mound since the nineteenth century, and the distribution of these features has only extended in recent years through small-scale excavation, aerial photography and geophysical survey. At first the main concern was with the relationship between Silbury Hill and the Roman road, but, as more was learnt about their surroundings, the site was interpreted as an important settlement, a mansio or even a small town. The coins discovered in the 1960s excavation at last had a wider context.

But what is the right framework for interpreting these features? Here Leary and Field have a distinct advantage, for their understanding of the Neolithic monument emphasises its distinctive siting beside a series of springs, one of which is the source of the Kennet, the principal tributary of the Thames. This leads them to question why a Roman town should have been located on damp, low-lying ground. For that reason the wells excavated in the nineteenth century cannot have played an entirely practical role. In fact the distribution of subsoil features focuses on the spring itself, rather than the dry land nearby. Could this have been the site of a sanctuary? An interesting comparison is the recently excavated site of Springhead in Kent. It included a wide variety of buildings, including a temple complex, and, as its name suggests, the structures were built around the source of an important river. It will take large-scale excavation to tell whether the same applies to the structures below Silbury Hill, but the suggestion is as important — and just as radical — as the authors' reinterpretation of the Neolithic monument. That is why it should be of interest to readers of Britannia.