This handsome volume records 24 years of research on Hadrian's Wall by English Heritage and its predecessor, the Department of the Environment. Two sections by Alan Whitworth reproduce 166 watercolours painted in 1877–96 by J. Irwin Coates, an invaluable record of many of the lesser-known stretches of the Wall, and describe Charles Anderson's work on the consolidation of the Wall from the mid-1930s to 1974. The latter section also covers the controversy in 1958 when Jacquetta Hawkes' criticisms of the methods used to clear and restore the Wall resulted in a Commons debate. This is an important but uncomfortable episode in the often troubled history of the Wall as an archaeological monument and needs to be explored more fully.

The rest of the volume is devoted to the excavations of various elements on the line of the Wall and to work at the forts of Birdoswald and Bowness-on-Solway. Of particular interest are the sections through the Wall and Vallum. At Blackcarts the digging of the Wall ditch was defeated by the hardness of the whinstone, even though a little to the south-east the Vallum ditch, dug through the same type of stone, was fully formed. Here and at Appletree the marginal mound on the south lip of the Vallum ditch was examined. Wilmott concludes that it was ‘primary or near-primary’ and that it might have been planned as an original element of the Vallum, in addition to the two mounds which flank the ditch to the north and south. The previous view has been that it represents upcast from cleaning out the ditch, possibly as part of a limited restoration of the Vallum following its slighting in c. a.d. 140. There are problems with W.'s new interpretation — not least the symmetrical profiles of the mound which seem to relate to its position on the eroded lip of the ditch rather than to the original lip — but investigation of the marginal mound must now be a priority in future research on the Vallum.

Exploration of the earliest levels in the north-west part of the fort at Birdoswald suggests that the original fort, like its stone successor, had projected north of the Turf Wall. The so-called hiatus level, previously seen to mark a temporary halt in the construction of the stone fort, extended into this area. W. regards the hiatus as of later Hadrianic date but does not link it to any historical event. An attractive context is the advance into Scotland in c. a.d. 140. It would provide an explanation for the filling of the Vallum ditch and digging of the southern fort ditches which are datable to after c. a.d. 150 and could mark the completion of the fort after the withdrawal from Scotland. Completion of the fort in the north-west area was marked by the building of two barracks north of the aisled basilica found in W.'s earlier excavations. There was probably space for four barracks in the north-east part of the fort and more than enough space, as W. observes, for the additional four barracks in the southern part of the fort which would be required by a cohors milliaria or quingenaria equitata. The two excavated barracks were rebuilt in the Severan period, apparently with eight contubernia as originally. The later levels were poorly preserved, but enough survived to suggest that the barracks had been rebuilt in the later third century as a series of free-standing contubernia.

Further work was done in the area immediately south of the fort which had been explored by Simpson and Richmond in the 1930s. Their sequence of activities is reinterpreted by W., who argues that the fort ditches, filled in the late third century, were superseded, rather than preceded, by a series of timber buildings. Finds in this area of ‘Housesteads ware’, pottery made locally in the Frisian tradition, point to the presence of a German irregular unit, as at Housesteads, Vindolanda and Burgh-by-Sands. The accommodation of military units in unenclosed settlements is puzzling in a period when their adjacent forts were being maintained and their defences refurbished. Perhaps at Birdoswald the unit was actually accommodated within an annexe, the defences of which are beyond the excavated areas. That would explain why the ditches were dispensed with on the south side of the fort.

Rescue work at Bowness-on-Solway, concisely described by Paul Austen, established the exact size of the western terminal Wall-fort. His contribution rounds off a varied collection which will be of enduring value in the study of the Wall and also contains much of wider interest.