This volume results from a series of conferences held in 2010 to celebrate the centenary of the Roman Society, which fortuitously fell 1,600 years after the traditional end of Roman Britain. Although the otherwise diverse papers are united in their view that the date of a.d. 410 is either incorrect, over-emphasised or irrelevant, it forms a convenient hook on which to hang a stimulating discussion about the archaeological and textual evidence for late and post-Roman Britain.

Esmonde Cleary provides a useful overview of the disciplinary history of late Roman Britain, highlighting views which fluctuated between ideas of continuity and rupture between Roman and Early Medieval Britain (or ‘long’ and ‘short’ chronologies) that were united by the rejection of a text-based narrative as an explanatory framework. A number of the papers here revisit this ground and there is an occasional sense of déjà vu in the discussion of some of the material evidence, with the protagonists for a short chronology (e.g. Faulkner) accepting the evidence of coins and pottery which indicate a rapid early fifth-century collapse while others (e.g. Fitzpatrick-Matthews) argue that we may be missing evidence or misinterpreting that which we have. Chronology remains a major issue and is central to the papers of Fitzpatrick-Matthews (on small towns) and Gerrard (on pottery), both highlighting the problematic nature of the evidence, while Roberts (on the kingdom of Elmet) demonstrates the value of radiocarbon dating (still under-used for this period). Many of the papers (e.g. those of Swift, Cool, Gerrard, Dark and White) also note the continued use and importance of earlier material in the late and post-Roman periods. The long life of coins is demonstrated by Moorhead and Walton and by Guest, who argues convincingly that the extraordinary quantities of metalwork recovered from Britain represent post-Roman rather than late Roman behaviour. The sudden disappearance of coins and fine ware creates a major lacuna that archaeology has thus far failed to fill, although Cool argues that the material culture suggests that the fifth century can be seen as a continuation of what was happening in the fourth, with a move away from fine ware pottery and a continued emphasis on glass drinking vessels.

The western peripheries of Roman Britain form the focus of several papers (Dark, Petts, White, Knight and Evans), which present sometimes conflicting interpretations of the evidence. Dark and White revisit the idea of the South-West belonging to a late antique world connected to Constantinople and the Mediterranean, while Petts (to my mind more convincingly) uses textual, epigraphic and material evidence to make the case for strong ecclesiastical and secular connections between the South-West and northern Gaul, arguing that the long-distance links of the South-West have been over-emphasised. Knight and Evans focus on Wales and the importance of the rebellion of Magnus Maximus in a.d. 383, noting the intriguing appearance of Magnus Maximus in medieval Welsh legends and genealogies in the form of Macsen Wledig, although Evans concludes (contra White) that any continuities between late Roman and medieval Wales are later constructions and that elements of romanitas in medieval Wales were reintroduced in the ninth century and later.

Scotland and the North are discussed in papers by Birley and Hunter who both highlight the fluid situation of the fifth century, although Hunter argues that for north-east Scotland the key period was actually the third century. He further suggests that sites such as Traprain Law indicate that warlords were able to take advantage of the complex situation in the wider world to which they were connected. The remarkable long sequence at Vindolanda also continued into the fifth/sixth century, where an individual called Brigomaglos is attested, suggested by Birley to be a warlord based at the site (a situation possibly paralleling that at Birdoswald). This idea of fragmentation and different local responses to a changing situation comes through very strongly in many of the papers. As Collins and Breeze note in their paper on the late military situation, there were three or four separate armies in Britain who followed very different trajectories.

This picture of a myriad different ‘ends’ and ‘continuities’ parallels that which has emerged for other areas of the Roman Empire in recent years and in Britain the data generated through research, the PAS and developer-funded excavation have created a much more detailed picture than was the case even a decade ago. This important volume richly illustrates the quality of the data and the possibilities that arise from a more nuanced and combined treatment of material and textual evidence within the broader context of the Empire. However, as Millett notes in his concluding remarks, the ongoing problems of chronology continue to hinder further understanding of this key period of Britain's history.