The study of Scotland's later prehistory has long been dominated by considerations of its diverse settlement record. This has three principal components: upstanding monuments, ranging from unenclosed house sites to major fortifications, a cropmark record including broadly similar categories of sites in the agricultural lowlands, and, in its many water bodies, a suite of lake dwellings. These, known locally as crannogs, normally demonstrate excellent preservation conditions, but have often been studied, while not wholly in isolation, with limited reference to their wider ‘dry’ hinterlands. Research effort has also generally tended to favour both the brochs and other exotic monuments of the Atlantic coastlands and offshore archipelagos, and the remarkable site visibility of the heavily grazed ‘sheepscape’ of the eastern southern uplands and nearby agricultural lands.

This important study breaks the mould through its focus on the extreme south-west of the country and the integration of wet and dryland field archaeology on and around the margins of a single small lowland kettle-hole loch, located near Stranraer. The latter in particular is a central aim of the Scottish Wetland Archaeology Programme, of which the leading authors are key members. In all, five sites around the margins of the water body were examined by excavation to varying degrees. This was supplemented by off-site environmental survey within the locality. The known crannog within the loch, Cults Loch (CL) 1, normally wholly submerged, had been examined on a small scale and radiocarbon-dated during an earlier underwater project and was not examined further. The earlier work also identified fluctuating water levels at CL as a threat to organic preservation and this provided an impetus for the work reported here.

The major excavation was focused on a crannog of different character, CL3, which was set on a promontory marginal to the loch. Here the natural configuration of the terrain was enhanced with copious amounts of wood (primarily alder), pegged in place and designed to support the footings of round buildings that were, aside from their hearths and some basal features, rather ephemerally preserved as a by-product of the oscillations in the water table. Major dendrochronological research was deployed to tie site use into oak sequences established for neighbouring regions, and to make a pioneering foray into the (more problematic) use of alder, the timber dominantly represented. As with the detailed Bayesian study of the radiocarbon determinations (Chapter 6), the dating evidence points to construction and use extending for about a century and ending late in the fifth century BC, conforming to an episode of crannog use identified elsewhere in the country. Efforts to tease out the sequence of two juxtaposed round houses set on the crannog demonstrated the probable sequence but were not conclusive. In the second century BC, the approach causeway was refurbished, but there is no other sign of subsequent recourse to this site. Good onsite preservation means understanding is enhanced by studies of insects, seeds and so on.

CL4 and 5, both dryland sites, underwent proportionately less intensive invasive fieldwork, but this was informed by preliminary geophysical survey. CL4 is categorised as a promontory fort, which projected into the margin of the loch, and was cut off to landward by a series of significant ditches (largely unexcavated) and distinctly slighter palisade lines, sometimes described—fairly to this reviewer's mind—as ‘fences’. There may have been three episodes to their construction. Restricted evidence that a substantial bank with a stone footing to its outer face sat on the scarped promontory edge elsewhere is also presented. The occupation evidence recovered from this dryland site is limited. CL5 is a much more conventional cropmark small enclosure set back from the loch and defined by an arc of ditch and a double palisade that is a potential candidate for a narrow box rampart. Internal occupation consisted of a substantial roundhouse. Considerably later, following a lengthy abandonment, this was replaced by a further round house, distinguished by opposed east–west entrances and a timber-lined souterrain—the first such structure to have been dug in the region—set partially within it, such that the timber house wall must have been carried over the wooden roof of the underground feature. Bayesian modelling of the radiocarbon dates suggests that phase 1 of CL5 may have been the earliest CL site to be occupied (probably 645–450 BC), and that its second phase may have run to c. AD 20, following an abandonment of perhaps three centuries. The overall impression is of successive, relatively short-lived occupations or uses (CL3 is highlighted as a liminal location where permanent settlement seems unlikely), in a pattern familiar from some areas of temperate Europe where Iron Age site chronologies are easier to construct.

In general, few artefacts, primarily of a relatively prosaic nature, were recorded. Identifiable wooden items were also scarce, but included an ard set amongst the foundations of CL3 (with parallels at sites such as Milton Loch 1) and a distinctive wooden box. The material culture found is set into a wider context through the inclusion of a very useful inventory and discussion of all Iron Age finds known from the county. This demonstrates a range of external links, including to the Atlantic west.

Much of the volume's focus is on the Iron Age, but the wider environmental study takes a longer perspective due to disturbance through the inwash of later sediments and the absence of peat of that date. Early anthropogenic indications include an isolated pit with hazelnut shells within CL4, although more probably to be Late Mesolithic than the Neolithic period claimed here (p. 140). A single first-millennium AD range-finding radiocarbon date for the pits examined in CL6 provides a hint of later activity, perhaps related to the submerged crannog CL1, although the proposition that they may have been graves seems an optimistic reading of the surviving evidence.

In the range and detail of the evidence presented, this volume is a significant addition to Iron Age wetland archaeology, as it is to the re-assessment of the form, function and use of such lakeside locales for living, however transitory and impermanent they may have been. In its detailed consideration of the chronological relationships between the promontory crannog and the nearby dryland enclosed sites, it both complements and contrasts with other recent contributions to south-west Scotland's later prehistory seen in wider perspective. Production standards are generally high, although this mature reviewer would have preferred slightly fewer, slightly larger, graphics; the volume is priced fairly. The results of the follow-up SWAP project, at Black Loch of Myrton, will be eagerly awaited.