Cornwall's north coast is notoriously hazardous for mariners, with large stretches of the coastline where there is no shelter at all. It is not surprising, therefore, that where such havens exist they have provided a focus for activity since prehistoric times. The Camel and Hayle estuaries are two such, and the sheltered porth at the mouth of the river Gannel, where Newquay now lies, is another. The headland of Trevelgue lies directly to the north of this, and has seen occupation from the Mesolithic (not coastal then, of course) to the post-Roman periods. Eight ramparts of varying size cross the promontory, part of which is now almost a detached island.

The site has long attracted the interest of antiquaries: Borlase excavated a barrow here in the 1870s. When in the 1930s the West Cornwall Field Club sought a major field project, Trevelgue was an obvious choice. After various vicissitudes, C.K. Croft Andrew was invited to direct the excavations, the last of the great ‘public’ excavations of the 1930s. Sadly, the first season of 73 days' excavation, by paid workmen and enthusiastic volunteers, was also the last, as work ended in September 1939, never to restart, and he never managed to write up the work. However, CKCA's results were always recognised as important, and to this day represent the largest excavation of a south-western promontory fort. Although he died in 1981, the archive, finds, and, importantly, a photographic record of rare completeness all survived. That this volume can present such a remarkable account of the work, of great value to modern studies, is a tribute to CKCA's recording, as well as to the energy and perseverance of first English Heritage, and then the two authors.

CKCA excavated seven principal trenches, by hand, mostly sampling the ramparts but including a substantial open area excavation of ‘House 1’. Phases of Mesolithic, Neolithic, Early and Later Bronze Age activity have been recognised, but the majority of the ramparts have been dated to the Middle Iron Age, although one may be earlier, and one, Rampart 8, may conceivably be Roman in date. As with many other defended sites in South-West England, there is no definite Late Iron Age structural evidence, although it must be remembered that less than one per cent of the site here was excavated. The one major non-rampart structure to be excavated, House 1, was a large, probably non-domestic structure with a long and complex history, used in the fourth to first centuries b.c., but then, after a prolonged period as a disused ruin, scraped clean inside to have a subsequent life as an unroofed focus of either ceremonial or possibly exchange activity after c. a.d. 320. Thereafter, in the early 5th century, it seems to have been ceremonially backfilled or buried with a rich range of midden material representing all preceding phases, apparently gathered from across the headland, and including more than 80 Roman coins, reported here by R. Reece.

An eroding iron lode crosses the headland, and it appears that this was exploited in prehistory. CKCA's excavations recovered a substantial volume of slag and other smelting evidence, examined by D. Dungworth. A Middle to Late Iron Age date range is proposed, and Dungworth suggests that the evidence points to a series of small bowl furnaces, ‘well suited to smelting small quantities of rich ores’. Other metalworking is indicated by the presence of small crucibles. The ceramic assemblage from the site represents the largest collection of first-millennium b.c. material from Cornwall. H. Quinnell's report, with petrological analyses by R. Taylor, represents a major contribution to the study of South-Western ceramics from the Late Bronze Age onwards, in particular on the South Western Decorated forms.

This volume is an impressive piece of work. The fact that so much valuable evidence can be drawn from an excavation 70-years-old is a tribute to the meticulous recording of CKCA's daybooks and photographs, the survival of the complete archive and finds assemblage, and the intellect and industry of the present authors. English Heritage must be commended for initiating the re-examination of the archive and funding this report — although I do regret the fact that they did not take the opportunity also to instigate a small modern excavation of the still-eroding rampart sections to provide a calibration of the earlier data and palaeo-environmental study of the buried soils. The stratigraphic report painstakingly teased out by the authors is accompanied by a series of excellent specialist reports which further illuminate the work (in summary in the printed volume and in full on the CD), and the volume is handsomely produced by Cornwall Council, albeit with a paperback binding not quite suited to so stout a volume.