The sea (and to a lesser extent the major rivers) and their potential for transport and supply are fundamental to understanding Roman Scotland. The fleet was critical in campaigning and the Agricola mentions their role in the initial conquest of Scotland and their famous circumnavigation of the country, while the Severan campaigns were equally dependent on maritime support (Agricola 25, 29, 38; Breeze 1982 , 131-2). This is seen archaeologically in the placing of some temporary camps (such as Dun, near Montrose) to take advantage of maritime connections. Water transport was also critical during the occupations, as the quickest and most economical method of transporting bulk goods (Green 1986, 39-41). This is seen most clearly in the distribution of pottery (e.g. Gillam 1973; see section 5).
Despite this, knowledge of the maritime element of Roman Scotland is poorly developed. Ports are hypothesised at such sites as Camelon, Cramond and Bertha purely on the basis of their positions (Martin 1992; Tatton-Brown 1980), although in some cases, as at Camelon, the diversity of pottery from the site supports this attribution. The question of feasible ports is closely related to reconstructions of sea-levels, estuary lines and the navigability of rivers, all of which may have changed significantly since the Roman period but topics which have seen only limited work (Tipping and Tisdall 2005, 444-7). Roman period wrecks in Scottish waters are as yet unknown.
Detailed palaeoenvironmental reconstruction of potential port sites and their likely access to the open sea might assist in confirming their putative maritime role, and also assist in the identification of the Roman shoreline and thus further areas to target for fieldwork and the search for the sheltered water or deposits where Roman wrecks might survive).
Any opportunity to investigate a wreck site should be seized; this ties in with wider questions of the study of the maritime heritage, addressed in the ScARF Marine & Maritime document.