The most important find of late Roman table silver from Britannia comprises 28 items, with a combined weight of 26,540 g (more than 81 Roman pounds); it includes a Bacchic platter, two Bacchic plates and a niello platter, a covered bowl, six flanged bowls, two pedestalled plates, 13 spoons and a fluted dish. Since 1977 there has only been a small but thoughtful handbook by Kenneth Painter, so Hobbs is to be congratulated on the full and detailed publication of this important treasure. The new study contains specialist reports on the technical aspects of manufacture and decoration (an important contribution by Janet Lang and Michael Hughes), on the graffiti (Roger Tomlin) and on the hoard's burial (Jude Plouviez).

The history of discovery is set out in ch. 1: the silver was discovered during ploughing near Mildenhall, not far from a poorly documented Roman villa; it came to the British Museum in 1946 and its exact findspot remains unknown. In chs 2–11 the objects are described and detailed measurements as well as photos and drawings are provided; each one is discussed in relation to relevant comparanda of the third to fifth centuries. Excellent photos support the comments on technology and illustrate the figural and geometric designs. It is also commendable that traces of wear are described, and capacities of vessels calculated. Table 2 provides a concordance of all previous catalogues.

Each vessel type is compared with sequences of related profiles, roughly from the oldest to the youngest. This approach of splitting a whole into individual elements and examining these in detail can make ordering difficult. Some specific comments: plates were sometimes deformed post-deposition — for example the apparently unique (32) Theodosius-Missorium from Almendralejo was undoubtedly flattened during restoration work in the nineteenth century. Overall, it is demonstrated that the height of the footring steadily increases from the third century onwards (37). But this trend is not a mathematical law, as is shown by the large, heavy platters Sevso 1 und 4 (on the development of platters see now M. Martin, ‘Der Silberfund von 1628 und das Tafelsilber des 5. Jahrhunderts’, in A. Kaufmann-Heinimann and M. Martin (eds), Die Apostelkanne und das Tafelsilber im Hortfund von 1628. Trierer Silberschätze des 5. Jahrhunderts (2017), 252–67, published since H.'s volume). For the seemingly unparalleled bowls 9 and 10 (155) it is worth reconsidering the decoration of the Water Newton unguentarium (K. Painter, The Water Newton Early Christian Silver (1977), cat. 5). Bowl 11 from the third century is a rare example of a vessel used for a considerable time (162 and 266, see A. Kaufmann-Heinimann in M. Guggisberg et al. (eds), Der spätrömische Silberschatz von Kaiseraugst. Die neuen Funde (2003), 134). That even the late Roman cover (172; the form of which is not comparable to bowls for imperial largitio) was not made specifically for the bowl is indicated by the fact that in contrast to all other silver bowls of this form the rim curves inwards (pl. 280).

Maps such as pls 48, 128 and 328 do not show the distribution of vessels, but the distribution of contexts, mainly hoards in which such vessels survive, a point made on p. 37. Equally, the dating evidence relates to hoards and contexts, not vessel types. There are few fixed points for chronologies; the issues surrounding the fifth century are not discussed in detail.

Chs 14 to 16 offer an interesting synthesis, first exploiting quantitative data, as has been done previously by the author (R. Hobbs, Late Roman Precious Metal Deposits c. AD 200–700 (2006)). Without epigraphic evidence, the location of silver workshops remains ‘difficult to answer and a matter of speculation’ (272). The measurements of the punches employed to make beaded rims suggested by the author as a possible way of identifying workshops (e.g. 264) — analogous to the use of die stamps — does not seem practical in the absence of other information about the vessels. Ch. 16 usefully summarises written and visual sources for the use of silver plate, with charming quotes from the Anthologia Latina (274). I am not convinced, however, by his interpretation of vessel functions, which are adopted from Hudson (‘Changing places: the archaeology of the Roman “convivium”’, Amercian Journal of Archaeology 114 (2010), 663–95). Interpretations based on ethnographic studies and finds from prehistoric Iran seem to me to be based on theoretical considerations and speculation.

H. has published a fundamental and stimulating study of a very important find, which will give rise to both praise and debate. He is to be thanked for the careful publication of the Mildenhall Treasure, which will enrich future research on late Roman silver in Britain and beyond.