These two volumes report on excavations immediately outside the legionary fortress at Chester. That on Gorse Stacks covers a single site dug in 2006 and 2008 in the northern zone, where almost all the Roman activity was confined to the later first and second centuries. A series of large pits had been dug to quarry sandstone and had then been filled with rubbish including large groups of pottery and other finds which apparently included very little material later than c. a.d. 120. The stone from the pits was of poor quality, though it was also quarried in the sixteenth century; it seems to have been extracted as rubble for use in foundations and wall cores. The volume by Ward et al. covers work undertaken between 1964 and 1989 in the western and northern zones. Consequently it is far more wide-ranging, describing some ten sites (reports by Ward, Mason and McPeake) and then presenting an exercise in artefact analysis which includes for comparison a survey of the objects from the burials at the Infirmary Field, mainly excavated between 1912 and 1917 (Heke); the final sections by Carrington are a valuable overview and gazetteer of all the evidence for the extramural settlements down to the end of 2009 and an essay which explores the economy and society of Chester and its regional setting.

Perhaps the most important excavation was at Greyfriars' Court, west of the fortress, where Trajanic buildings of timber construction were succeeded before the end of the second century by two phases of substantial stone buildings, the first of them representing a complete replanning of the area. The presence of buildings on this side of the fortress which were not for ordinary domestic use was confirmed by trenches at Greyfriars' House, which uncovered a room with an area of at least 57 sq m and a hypocaust combining channels and pilae, and at Sedan House where more of the Watergate Street baths, first seen in 1778/9, was explored. All these excavations were in the so-called clear zone, a strip of land about 150 m wide encompassing the defences of the fortress, which was either occupied by buildings of an official nature, including the amphitheatre, or at first undeveloped. Carrington shows that this zone, where not built over, was soon used for various industrial processes, including quarrying, as at Gorse Stacks, and pottery manufacture, as suggested by a few wasters from that site. At some stage the western part of the zone was enclosed by the ‘quay wall’, convincingly reinterpreted by Mason in 2001 (see now Roman Chester. Fortress at the Edge of the World (2012), 109–10) as a defensive wall forming an annexe to the fortress, an idea further developed by Ward in the volume reviewed here.

The size of the extramural settlement is compared by Carrington with that at York which, even excluding the walled area of the colonia, is much larger than at Chester. He examines the possible functions of the extramural areas and is surely right in placing the emphasis at Chester on servicing the fortress and the network of supply routes which spread out from it, whereas at York the size and status of the colonia might suggest that it was not entirely dependent on the legion. At Chester, therefore, the history of the fortress might be expected to dictate, more or less entirely, the development of the areas beyond its walls. Carrington makes much of the ‘large-scale absence’ of the legion after c. a.d. 120, an idea current for the last two or three decades, which he expresses perhaps in its most extreme form, stating that the fortress was only held in strength between a.d. 90–120 and 210–250. This idea results partly from the assumption that most of the legion would have been absent for the whole duration of building work on the two Walls and their forts; it seems more likely that the cohorts returned to their main base after the building seasons rather than spending winters on the frontier in the unnecessary discomfort of makeshift accommodation.

A larger assumption seems to be that the Roman authorities would have expected that the fortress would normally be occupied in strength and that, when events conspired to make this the exception rather than the rule, many buildings were left to go to rack and ruin. In the case of long-established fortresses, the expectation was surely the reverse: that normally much of the legion would be posted elsewhere — in its province, on the frontiers or serving as vexillations in other provinces. Fortresses served as headquarters and reserve bases, and to serve those purposes their buildings would need to have been maintained in reasonable condition. At Chester some of the evidence proposed for the dereliction of parts of the fortress in the second century is questionable, as indeed is the apparent scarcity of datable finds of this period when compared to the quantities from the Flavian-Trajanic period. The only finds which have been surveyed from Chester as a whole are the coins which, as Carrington shows, have a rate of loss above the mean for British military sites throughout the second century.

In the Gorse Stacks report, the supposed ‘large-scale absence’ of the legion has clearly influenced the interpretation of the results. The artefact analysis by Cool concludes that at least some of the material from the filling of the quarry pits ‘was undoubtedly derived from within the fortress’, more specifically that it came from middens that had accumulated in the fortress up until the a.d. 120s but which were not cleared away until the late second century. Timby is justifiably doubtful about the source of the pottery found in the pits, admitting that the differences in the types used by those inside and outside the fortress ‘may not be very clear cut’. The samian (reported on by Wild) and coarse wares are large closely dated assemblages, the importance of which is brought out well by the two authors of the pottery reports. Equally thorough are the finds reports in the other volume. Particular attention is paid to the small finds (Heke) which are compared with assemblages from the eastern canabae at Caerleon and from the fortress and colonia at York. There are stark differences among some of the categories: hairpins, for example, made up 27.5 per cent of the personal items from the York colonia, 12 per cent from the York fortress and 2 per cent from the Chester canabae sites studied by Heke. To what extent these sorts of results shed any further light on the character of the occupation already established by the structural remains is at present unclear. This is clearly work in progress, which would be helped, if only the evidence were more plentiful, by comparing like with like — baths inside and outside the fortress, barracks and strip-houses, larger houses in the extramural settlement and centurions' quarters, and so on — rather than miscellaneous areas inside and outside the fortresses, each with ranges of use and changes of use through the centuries.

Overall, these are excellent excavation reports, and Carrington's contributions provide the essential theoretical framework for future excavations in the extramural areas. The report on the excavations in 1964 to 1989 is mainly by former staff of the Grosvenor Museum and Chester Archaeology, the latter now succeeded by Cheshire West and Chester Historic Environment Service. The Gorse Stacks excavations were carried out by Birmingham Archaeology under the provisions of PPG 16, and it is encouraging that the result has been a highly competent report promptly published. But for the present planning protocols to continue to work effectively to the benefit of complex sites of truly international significance such as Chester, they need to be underpinned by the detailed academic knowledge provided by locally based teams. It is widely known that changes are planned in the status of the archaeological services at Chester and, as at so many other places in England, there is no guarantee that the accumulated expertise of several generations will be safeguarded and developed.