Lankhills was one of the group of excavations on fourth-century urban cemeteries that in the 1960s brought funerary archaeology to the forefront of Romano-British studies. Its 1979 publication by Giles Clarke not only set standards for the publication of such sites, but was also ground-breaking in arguing that analysis of burial rite could enable the identification of ‘incomers’ to the cemetery, particularly in this case a group of burials dating from a.d. 350 onwards and originating in the region of the central Danube, his ‘Pannonian’ group. From 2000–5 Oxford Archaeology (OA) excavated an area contiguous with the northern side of the 1967–72 site, adding another 329 burials to the 451 of Clarke's work. Like the latter, the Oxford work has been promptly published to a very high standard in the presentation of a wide range of evidence, in the excellent plans, object drawings and photographs, and especially in the full, clear and judicious texts. This report draws heavily on Clarke's, and the discussions (especially ch. 7) and the plans synthesise the information from the two campaigns of excavation, maximising the yield of the report. This is a first-rate research publication arising out of a commercial project.

Broadly speaking, the OA results conformed to those of 1967–72 in terms of the development of the cemetery, the types of burial rite and of material culture, and the dating, though with more importance for cremation. The burials were overwhelmingly inhumations, with parity of males and females though few juveniles, starting in the earlier fourth century. For the first half of that century the dominant rite was unfurnished, coffined inhumation, some burials having simple offerings such as a pot or hobnailed shoes. The discussion emphasises that at all times unfurnished burial was the norm, against which the ‘special’ rites need to be set rather than being decontextualised. It also shows that in all rites the offerings were of low intrinsic value, even in the more elaborate burials; in the latter did they stand proxy for richer items used in life? For three topics in particular this volume was able to draw on resources not available to the 1979 report. First, comprehensive osteological analysis; second, the use of stable isotope analysis on the human bone; third, much fuller and more nuanced appreciation of topics such as ‘identity’ and ‘ethnicity’.

Amongst the more elaborate burials were a dozen or so that broadly matched Clarke's criteria for his ‘Pannonian’ rite, albeit the objects were of types made in or standard for Britain. Some of these skeletons were subject to stable isotope analysis along with samples from ‘indigenous’ burial rites to try to ascertain whether the ‘Pannonian’ group was of non-British origin; and whether it included local-origin people (women? children?). There was, of course, a mis-match between burial rite and isotopic signature; only one burial had a ‘colder’ signature suggesting an origin in central/eastern Europe but it did not have the full ‘Pannonian’ rite; many that did have that rite had ‘local’ signatures. Most of the non-British signatures were ‘warmer’ suggesting origins further to the south (the ‘warmest’ of which had two North African unguent flasks) yet were of ‘indigenous’ burial rite. At one level this makes identification of ‘intrusive’ rites and individuals much more complex, at another it shows an unexpectedly cosmopolitan parent population for Lankhills. The use of a wider range of burial rites at Lankhills was characteristic of the second half of the fourth century, though Lankhills remains exceptional when compared with other fourth-century cemeteries at Winchester, and indeed with other fourth-century cemeteries in Britain (save Scorton by Catterick). One future avenue of research will be comparison with continental burial rites and material where again a minority of ‘special’ rites lies alongside an unfurnished majority at a variety of site types. The second half of the fourth century also saw major changes within the walls of Winchester, so the town may have been changing its nature, though whether and how the changes in burial relate to this is unclear. Burial at Lankhills seems to cease abruptly in the earlier part of the fifth century.

This report fully deserves to stand alongside Clarke's 1979 publication, and will likewise continue to be a source of evidence and ideas for a generation and more.