Advocates of the dictum ‘Geography is destiny’ might well cite the location of Roman London in support of their case. Situated at what appears to be the lowest bridgeable crossing-point on the tidal Thames, the area that was to become Londinium provided relatively high, dry and flat ground on the north bank, allied to potentially good harbouring facilities. Gravel islands on the south bank allowed roads to be brought up to the Thames, cross the river and continue into the interior.

That the transpontine settlement, with an urban area envisaged as perhaps 20–25 per cent of that on the north bank, is likely to have been an ‘integral’ part of Londinium, rather than a ‘suburb’, is accepted in this important volume. Whether the evidence discussed weakens or strengthens the hypothesis that Londonium is more official than civic is a matter for readers to assess.

This volume brings together 41 previously unpublished excavations that were carried out between 1973 and 1991 and draws on the findings from others already published, in an effort to provide a synthetic account of Roman Southwark. It reports on investigations into Southwark's early topography which reveal that much of the settlement lies on its ‘north’ and ‘south’ islands, probably narrowly separated from the more expansive gravel flood-plain to the south. This natural topography is featured on a series of well-illustrated and thought-provoking period maps.

After an introduction (ch. 1), further chapters examine the settlement's topography and chronology (ch. 2), and address a range of topics relating to infrastructure, economy and religion, as well as to specific periods (ch. 3). Some conclusions are offered in ch. 4, followed by a series of appendices on finds and environmental evidence (ch. 5) and a gazetteer of the 100 sites included (ch. 6).

Excavations on the highways have revealed much of the course of Road 1, essentially the northern continuation of Watling Street and Stane Street approaching the bridgehead. Stane Street itself remains elusive in Southwark: although it is shown as joining Watling Street at the conventional position on the edge of the flood plain gravels, there may still be a possibility that the line is straighter and the junction further north, on the south island itself. Indeed, when the distribution of the sites investigated is plotted, it is apparent how limited the archaeological information still is.

In terms of chronology, there is now clear evidence for structures in Southwark preceding the Boudican revolt. This is especially so close to Road 1 on the north island where granaries were consumed by a fire at the time of the uprising. There is evidence from military equipment of an early army presence, perhaps for purposes more than just infrastructure projects. Examination of building material suggests that pre-Boudican masonry buildings were standing on as yet unlocated sites. The excavations have clearly demonstrated that ‘official’ buildings are likely to have been an important constituent of Roman Southwark through much of its history. These include the presumed ‘mansio’ (Site 45), between Roads 1 and 2 on the north island, and the range of buildings (Site 3) close to its north-western edge which produced a military inscription and CL BR tiles.

Interesting questions about economic activity in Southwark arise from Site 1 where more than 60 iron-smithing and bronze-working hearths and evidence for the manufacture of tools, nails and fittings were uncovered. The production is described as ‘more intensive and longer lasting’ than from any other known site in Roman London, perhaps suggesting similarities with the ‘industrial’ Walbrook valley across the Thames. It may also suggest a role for river barges, like the one found in the ‘Guys channel’ in 1958 (Site 8).

The volume contains an illuminating series of essays on Roman Southwark, related closely to the sites and finds reported on. The survey of evidence for religion and ritual is outstanding; thoughtful, thorough and well argued. Some essays, perhaps because of problems inherent in the evidence, are less convincing. For example, in terms of the geography, it may well be that Southwark had, as suggested here, no Thames side ‘port assemblages’ that link it functionally with the developed series of waterfronts across the Thames. Yet when much of the medieval Thames frontage was destroyed by massive thirteenth-century erosions, the evidence for much of the Roman river edge would have gone too, assuming that it had survived until then.

Change or continuity through time can be difficult to ascertain. Though the significance of Macphail's work on ‘dark earth’ is commented on here, assertions made as to the differences between early and later Roman Southwark, particularly those that minimize the size and duration of the latter, need to be examined cautiously. That might include the suggestion of a late Roman contraction and a withdrawal of the urban core towards the security of the bridgehead, particularly if arguments advanced elsewhere for the bridge being down by the middle of the fourth century are true. Ditches near the bridgehead may indicate that this did happen, but perhaps very late in the period. Burials in the Roman tradition clearly do occur within the urban core where buildings once stood, but most of the graves are without goods and how many were interred prior to c. a.d. 400 is unclear.

But these are small reservations, perhaps pointing to the need for more work on the ground as development occurs, particularly in the areas that have as yet received little attention: this volume is an invaluable contribution towards the study of Roman Southwark and a testimony to the skills and dedication of its major authors and many contributors.