This substantial volume is the latest product of the University of Reading's ongoing programme of excavation of a large sample of Insula IX at Silchester; the processes of post-excavation analysis and reporting have been running in tandem with the fieldwork and reports already produced include an account of the Late Roman sequence of the site (M. Fulford, A. Clarke and H. Eckardt, Life and Labour in Late Roman Silchester Excavations in Insula IX since 1997, Britannia Monograph 22 (2006)), reviewed in Britannia (40 (2009), 358–9). In addition, the sequence relating to the principal eccentrically aligned buildings (‘House 1’) has also been published in Internet Archaeology (in 2007). This allowed provision of more detailed data, but of course requires an additional payment (and assumes a reliable internet connection) and is admittedly (e.g. 9) already superseded in some important points of interpretation. The benefits of supplementary digital reporting are therefore mixed, but with regard to conventional reporting the overall timescale of the project makes the staged approach highly desirable; the obvious disadvantages of publication before completion of excavation are acknowledged by the authors, but are surely more than outweighed by the benefits — in particular the dissemination of new ideas about Romano-British urban development as exemplified by Silchester, and the presentation of scientific approaches to deposit analysis, intensively applied here.

Presenting the middle parts of a sequence is more difficult than dealing with its beginning and end, however, and the problem is exacerbated here by the scarcity of extensive surviving deposits which allow clear stratigraphic links to be made between spatially distinct areas and features; the scheme of phasing therefore relies principally on the sequences within individual buildings and the dating of scattered pits. The definition of beginning and end points is inevitably flexible. The concise narrative (9–52) describes activity assigned to Periods 3 (c. a.d. 125/150–200) and 4 (c. a.d. 200–250/300). The principal elements are the development of the sequence of buildings aligned obliquely to the street grid (an alignment already established) as buildings MB 1, MB 2 and MRTB 1, with further buildings MRTB 2 and 3 adjacent to and aligned upon the north–south street, all in Period 3. In Period 4 the oblique alignment was maintained by a single structure, MB 3, while MRTB 2 and 3 were also replaced by structures on the same alignment as their predecessors. Most of the plans and photographs which accompany the narrative are in colour (in contrast to Life and Labour), to its considerable benefit. Colour is also used in the two following chapters which deal, perhaps unexpectedly, with the geochemistry of the ‘House 1’ sequence in Periods 3 and 4 and microscopic perspectives on the use of MRTB 1 and its predecessor specifically in Period 3.

As Banerjee's introduction to the latter chapter rightly points out, ‘the use of internal space is an under-studied area of Roman archaeology’ (63). A key result of this work is the demonstration of non-ferrous metalworking (including gold and silver) in both periods. Equally striking is evidence for skinning of dogs (and badgers) in Period 4, amongst indications of more predictable economic activities. A variety of finds and environmental evidence is presented in chs 5–17, some supplemented by further material in appendices. Ch. 18 (by Eckardt) gives a detailed analysis of the pits in context, complementing a similar analysis in Life and Labour, and is followed by a short chapter (Matthews) on ‘visualising’ (rather than ‘reconstructing’) the buildings, reflecting the importance of such work not only in presenting the evidence to a wide public in accessible form, but also its value in making excavators think through all aspects of the site record — nevertheless, such visualisations represent the possible, rather than the actual. The shabby appearance of some of the buildings seems quite plausible to this reviewer, but some of the overall building heights seem uncomfortably low; a nice touch is the appearance (340) of one of the principal authors lugging a (hopefully empty!) Dressel 20 amphora.

The closing discussion neatly integrates the various strands of evidence to present a compelling view of the site — pointing up both mundane and less usual aspects of the range of practices seen there, within a framework which envisages substantial continuity of family-based landholding from the as yet incompletely excavated early Roman phases. As with visualisation, some points will appear more convincing than others — for example does the addition of a second storey to the buildings double the number of occupants of this part of Insula IX (e.g. 341)? Overall, however, what is most satisfying is the rounded picture which emerges (including the fact that interpretative challenges are admitted rather than sidelined), and we can look forward to the final instalment of this important series of reports.