This volume of fourteen papers is the result of a conference held in Newcastle in 2008, and the editors are to be praised for the production of this well-illustrated and well-presented volume in such a short time. The conference aimed to ‘elucidate the life of the fourth-century limitanei of Britain through their material culture’. Both the chronological and geographical scope are broadly conceived. Many authors exploit assemblages that have been recently excavated or re-assessed; the fourth century is understood as ‘the long fourth century, i.e. the later third to early fifth century’. Indeed, a deliberate effort is made to link the discussion to the post-Roman period (e.g. contributions by O'Brian and Roberts). The equally broadly defined frontier zone goes well beyond Hadrian's Wall and essentially encompasses the area between the Humber-Mersey and the Forth-Clyde line.

The papers share a number of common themes, in particular a desire to highlight that while the nature of finds changes, the fourth century does not necessarily represent a period of decline. Both bulk finds and small finds are covered, as is the environmental evidence (Huntley and Stallibrass). The wider context is provided by Hassall and O'Brien's discussion of epigraphic evidence and by Wilmott's summary of the changing design of structural remains, which he relates to changes in organisation and supply.

Bidwell and Croom provide a useful summary of pottery supply and use on Hadrian's Wall, discussing changing production sites, the relevant dating evidence, and whether the apparent simplification of supply in the late fourth century is due to military contracts, successful competition or wider economic factors. Price summarises the use of late Roman glass on Hadrian's Wall and beyond, noting its rarity in the fourth century even on important sites such as Carlisle and Corbridge. Both chapters illustrate how changes in form relate to changing social practices, with, for example, a wide range of specialised pottery types being replaced with cooking pots and jars. Brickstock attempts to establish levels of pay in the later Roman period, and concludes that while third-century soldiers had a very low income, imperial donatives and the annona combined to give their fourth-century equivalents a decent standard of living.

Several other papers deal with small finds, mainly relating to personal appearance, and how these relate to the changing social structure of the province. For example, Cool illustrates how first-century military assemblages are dominated by items of armour, while in the fourth century women are much more conspicuous. Allason-Jones provides an overview of personal appearance, while Coulston (on military equipment) and Collins (on brooches) provide detailed case studies, highlighting how northern finds relate to southern British and indeed empire-wide patterns in production and use. Both, like Hunter, also include a very useful corpus of the material discussed and all these contributions highlight the impact of the Portable Antiquities Scheme. Careful mapping can highlight regional differences, such as differences between the eastern and western parts of Hadrian's Wall (see papers by Cool, Coulston, Collins, Bidwell and Croom, and Roberts).

Hunter explores the regional and chronological patterns in the northern material, and argues that late Roman imports reached the area through trade, looting or as diplomatic gifts. He also discusses changes in the indigenous material culture, where several objects (such as ring-headed pins, massive terrets and knobbed spearbutts) show links with Irish and English material that may relate to the recruitment of mercenaries and more generally to the movement of people.

The two final papers (O'Brien and Roberts) trace the use of the Wall and its hinterland into the early medieval period, exploring the transition from late Roman commanders to warrior élites through the epigraphic and burial evidence, and through the analysis of maps and place-names.

As a whole, the volume provides a useful insight into an often neglected period, and highlights that, as so often, we are not dealing with a generalised pattern but with a patchwork of localised responses.