The Channel Tunnel Rail Link (CTRL; now known as High Speed 1) was the largest linear infrastructure project in Britain since the heyday of motorway construction in the late 1960s and 70s. It was also the first major scheme to be designed and implemented with regard to the regulations for Environmental Impact Assessment introduced by the European Union in 1985. These required that the impact of the railway upon archaeological remains be established, and that appropriate mitigation (either preservation or recording) be implemented. CTRL was therefore at the forefront of the development of the methodologies for integrating archaeology as part of the construction process, and the contrast with the desultory approach during the construction of the M40 motorway in the late 1980s could not be starker. The work of reconnaissance evaluation and subsequent excavation and watching-brief took place over a decade from the early 1990s and served to double the number of sites on the route compared with those known before work started.

Three Romano-British sites scheduled as ancient monuments (the religious sanctuary and roadside settlement at Springhead and the villas at Thurnham and Northfleet) were impacted upon by the railway. CTRL therefore provided the opportunity for the large-scale excavation of types of sites which have been comparatively little investigated by developer archaeology in the last couple of decades. More generally these publications allow us to review just what can be achieved by well-designed, well-funded and well-executed investigations in advance of construction.

The fieldwork reported here was undertaken by Oxford Archaeology and Wessex Archaeology and they have commendably collaborated on the publication of their results (although the joins where the Springhead and Northfleet reports have been spliced together are sometimes evident). A consistent and unified approach to reporting was not adopted across the scheme due to the longevity of the archaeological fieldwork. These volumes only cover work in Kent: investigations in Greater London and Essex are published elsewhere (including some interesting early Roman inhumation burials from West Thurrock). Within Kent the scheme was divided into two sections for construction purposes. Section 1 ran from the Channel coast to just south of Springhead. The approach adopted here was for all the detailed reporting of site stratigraphy, finds and environmental data to be placed in grey literature on the internet via the Archaeology Data Service (doi: 10.5284/1000230), with On Track as a book-form discussion and contextualisation of the results. For the prehistoric and medieval periods these were written by university academics not involved in the fieldwork; for the Roman period by Paul Booth. For Section 2, which comprises the Ebbsfleet Valley sites of Springhead and Northfleet, a more conventional approach has been adopted, with the evidence reported in a high level of detail in four lavishly illustrated volumes. The Springhead and Northfleet reports start with very detailed, and frankly quite dry, stratigraphic accounts followed by a wide-ranging and thought-provoking discussion chapter. It would help the reader if the volumes had begun with a much more extensive section outlining what was already known about each site, the extent of previous investigations and how these relate to the CTRL work. Indeed some readers might find it profitable to read the discussion first before wading into the site narratives.

On Track is an interesting approach, but of course one cannot contextualise something without providing at least some description of it first. The volume therefore briefly describes the main findings, and those who do not wish to interrogate the evidence deeply may well find that it gives them all they need for an overview. In contrast the Ebbsfleet Valley volumes present the evidence in a length and detail that is rarely possible these days. Literally no stone is left unturned or unreported, and specialists in whatever field will surely find all that they require between the covers. While the Section 2 approach will probably be more familiar and comfortable to most readers, that for Section 1 is the one which is likely to become more prevalent given the ever-increasing number of analyses which can be applied to archaeological data leading to ever longer, often highly technical, volumes which sell in very small quantities.

It might have been thought that a project such as this would make a major contribution to knowledge of non-villa rural settlements, but in the event only four sites lay sufficiently within the route for intelligible layouts to be recovered (and even here building plans were scarce). Surprisingly, given the amount of development in Kent in recent times, knowledge of this type of site remains slight and so the CTRL work is an important first step in characterising rural settlement in the county. Considerable effort was also devoted to landscape reconstruction, although the results were variable. While it would appear that the landscape was essentially open in the Roman period it has not proved possible to draw out many conclusions on the balance between woodland, arable and pasture. The highlights of the project are therefore the two villas and Springhead (including the Pepper Hill cemetery 500 m to the south; another slice through this rich cemetery containing bustum burials and early Roman inhumations was subsequently excavated during the upgrading of the A2 trunk road and has now been published by Oxford Archaeology).

At Springhead the ritual landscape surrounding the spring has been elucidated, with a trackway leading across the hillside to terraces, interestingly interpreted as places where spectators could gather to view rituals taking place below. The road layout as now revealed at Springhead is puzzling, with a pronounced dog-leg in ‘Watling Street’. The roadside properties were aligned along a linear stretch which comprises both what we would term ‘Watling Street’ and also a seemingly contemporary branch road heading northwards to Northfleet and an inlet of the Ebbsfleet, a tributary of the Thames. It almost looks as if the southern stretch of Watling Street and the branch road might have been the primary route, in which case Watling Street would have originally terminated at Northfleet, with the westward extension towards London which creates the dog-leg effect a second stage in the development of the road system. The characterisation of the activities taking place in the roadside properties and backlands is insightful and important. The authors wisely refrain from judgement on the question of an early military presence at Springhead, although if fragments of ditch do define a fort then this must have been very short lived as Watling Street cuts obliquely across it.

The two villas demonstrate interesting sequences of development. At Thurnham a Late Iron Age enclosure continued in use until c. a.d. 60/70 when a stone building (termed the proto villa) was constructed. This was replaced by a larger masonry house in the early second century. At Northfleet the main house apparently lay outside of the excavation area (it is known from limited recording in 1909–11). Rather the work revealed two masonry aisled buildings and a small bath-house adjacent to a quay on the Ebbsfleet. The stratigraphic and environment data combine to demonstrate that the brewing of ale was a significant, if not indeed dominant, aspect of the villa economy and the authors are able to reconstruct one of the most complete pictures of brewing in Roman Britain. They suggest that the villa estate might have been set up under official control to supply the army, for which they seek support from a possible fragment of an armilla (a military decoration) and a seal-box in the shape of a beneficiarius lancehead, a symbol used by state officials in Upper Germany and Raetia. To this reviewer a private establishment set up to exploit the opportunities of supplying beer to London seems equally plausible.

One of the most striking aspects of the CTRL sites is the dominance of first- and second-century activity — the rural settlements examined appear to be in terminal decline by around the middle of the third century, and this is true for both villa and non-villa sites. Thus at the time following the construction of the town walls of Canterbury and the coastal Shore forts the evidence from CTRL is for the total abandonment of some sites and reduced occupation at others. Just what was happening in the Late Roman countryside in Kent is perplexing. Was there a retreat to a smaller number of sites caused by factors such as changes in agricultural practice and land-tenure; exhaustion of the agricultural potential of the land; or a reduced rural population? Or did the decline in the Wealden iron industry have a profound effect on the wider agricultural landscape of Kent? The contrast with the affluent Late Roman villa landscapes of South-West Britain could not be starker.

Many millions of man hours have been expended on archaeology on CTRL, and these and other publications are a testament to just what has been learnt (and obviously also for periods other than the Roman one). Oxford and Wessex Archaeology are to be congratulated on co-operating on the production of these volumes which amply demonstrate the knowledge dividend that accrues from high-quality fieldwork and analysis. It is hard to believe that anyone who reads these volumes will doubt that important research can be achieved in a development framework. Their publication is also timely given that plans are now being developed for the next high speed railway between London and the Midlands, and ultimately the North. What has been learnt here should inform the design of the archaeological work to come.