Many sensational discoveries were made in the Hadrian’s Wall zone during the 1860s and 1870s, not least seventeen large altars found in pits outside the fort at Maryport on the Cumbrian coast. In the earliest accounts, their burials were explained as respectful acts, protecting them from harm when the fort had been temporarily abandoned. Later, they were seen as a series of ceremonial disposals, when vows to Jupiter were renewed annually and altars dedicated in the preceding years were removed. In 1880, ten years after the altars were found, two temples were excavated on a site about 100m south-west of the pits. Questions remained about the burial of the altars and their possible connection with the temples, especially after convincing arguments against the supposed annual disposals emerged. The project reported on here tried to resolve these problems. Excavating sites previously investigated by antiquarians is no easy matter, and the careful approach and technical virtuosity displayed in this report are admirable. Equally, much vital information is contained in the specialist contributions, and of particular value is the study of the altars from the point of a view of a mason by our late Fellow, Peter Hill.

Ditch systems were laid out in the early second century, and the Circular Building, the earlier of the two temples, followed soon after. Set centrally in a ditched enclosure, it was aligned on the point of sunrise at the summer solstice. Ian Haynes, credited with the overall discussion, considers that there might have been a connection with the cult of Sol, while not excluding the identification of the building as a tropaeum. Its modest specifications are seen as no obstacle to the second of these interpretations, but a victory monument surely ought to have involved the lavish use of manpower and materials. The second building is dated to no earlier than the mid-third century. The building of a classical temple with a tetrastyle portico at such a late date is exceptional, let alone in the fourth century, which the authors argue might be possible. Its dedication is uncertain, but, like its earlier neighbour still standing, the new building was aligned on the summer solstice at sunrise. Discussion by Haynes of the implications of this late temple-building is of great interest. It occurred in a period when the settlements outside many forts in the Hadrian’s Wall zone were being abandoned. We should now expect at least some of their cult areas to have remained in use for rather longer.

The axes on which the two temples lay ran across the slightly higher ground where the altars were found. During the earlier Roman period, the area was occupied by a polygonal enclosure defined by a small ditch. In its interior were a scatter of postholes and other slight indications of structures, but nothing to suggest the presence of substantial buildings. The later Roman period saw the substitution of a circular ditch, again defined by a ditch of modest size. The near-complete excavation of these two successive enclosures together with the area in front of the temples showed beyond much doubt that the altars had originally been erected elsewhere.

‘Arguably the most important aspect of the excavations’ (p 89) was the recovery of evidence which persuaded the authors that the altars deposited in the pits were merely part of the packing around large timber uprights; there was no reason to think that they had been disposed of ceremonially. The rows of pits seemed to represent a building or buildings dating to no earlier than the end of the fourth century. The authors are cautious about this interpretation because there are no real parallels to the fragmentary building plan and its eccentric construction technique. The reader will wonder if there is an alternative, but the organisation of the report makes it difficult to pursue other ideas. Missing is a summary that sets out and evaluates the evidence for the sequence of late activities and their date. A crucial example is the date of the pits, where it should have been emphasised that their late Roman date depends on a single stratigraphical relationship. The pits, or at least those of Row F, are said to have been cut through a ring ditch that contained late third- or fourth-century pottery and one sherd no earlier than c ad 370; this single sherd could be up to a century later than the other pottery in the ditch. The crucial relationship was visible in one pit of Row F, which apparently preserved enough of its original filling to establish that it had not been cut by the ditch, or at least that is what reference to an appendix suggests. A fuller description in the main text would have been helpful, for the only datable find from the original filling of any of the pits, apart from the altars themselves, was a Republican denarius, and there was no material suitable for radiocarbon dating.

There were perhaps as many as a dozen rows of pits, all running roughly from north-east to south-west. Few were exactly parallel to each other, the pits or post-settings in adjacent rows were rarely placed in opposite pairs, and the spacing of the pits was often irregular. The authors in their figure 220 try to make sense of this confusing tangle as a building, but have to exclude many of the pits from their plan on the grounds that they represent repairs or rebuildings. When simple methods requiring no more than stakes and lengths of cords are all that are needed to lay out the components of rectangular buildings, the irregularity of the supposed post-pits at Maryport is baffling. The meandering rows bring to mind the development of cemeteries, where new burials are positioned by eye in respect to the markers of earlier burials. A review is not the place to propose new interpretations, but there is still much to be said for older views of the pits as essentially the means of disposing of the altars respectfully, perhaps together with organic cult materials which have perished, for example wooden images and dedications.