This study has attempted to examine third and fourth century pottery supply in northern England together with other aspects of pottery assemblages in the region. The pottery kilns of the East Yorkshire industries have been characterised by neutron activation analysis which has proved reasonably successful in discriminating between them. Neutron activation analysis has also been utilised to attempt to check visually identified fabric groups and to help isolate other fabrics. This has been of varying success. Quantified data has been collected from 15 sites across the north and the limited published quantitative data have been utilised to examine the distribution, marketing and competition between fabric types in the region. Examination of functional variations through time between different types of site has also been undertaken as has that of variations in the quantity of finewares through time and between different types of site together with an attempt at quantifying decoration and examining trends in this. Pottery supply to the northern frontier area would seem to have been organised by different mechanisms in different periods. In the second century much of the pottery used on the frontier would appear to have been produced by the military themselves whilst in the third century and earlier fourth century free market mechanisms would seem to have operated, but in the late fourth-early fifth centuries some form of 'military contract' would appear to have taken over supply. Functional variations between different types of site have been identified with rural sites, turrets and Signal Stations sharing a major emphasis on the jar as the basic ceramic form and more complex settlement types having more diversified functional groups. The distribution of finewares also seems to be concentrated on more complex settlement types. It is apparent that there are consistent differences between the East Yorkshire region and the rest of the study area which may well reflect differences extending back into the Iron Age. Similarly there seem to be indications of some 'de-Romanisation' in late Roman assemblages but this does not develop in the fifth century, when nearly all the strands of evidence of Romano-British material culture disappear very rapidly.

This study has attempted to examine third and fourth century pottery supply in northern England together with other aspects of pottery assemblages in the region. The pottery kilns of the East Yorkshire industries have been characterised by neutron activation analysis which has proved reasonably successful in discriminating between them. Neutron activation analysis has also been utilised to attempt to check visually identified fabric groups and to help isolate other fabrics. This has been of varying success. Quantified data has been collected from 15 sites across the north and the limited published quantitative data have been utilised to examine the distribution, marketing and competition between fabric types in the region. Examination of functional variations through time between different types of site has also been undertaken as has that of variations in the quantity of finewares through time and between different types of site together with an attempt at quantifying decoration and examining trends in this. Pottery supply to the northern frontier area would seem to have been organised by different mechanisms in different periods. In the second century much of the pottery used on the frontier would appear to have been produced by the military themselves whilst in the third century and earlier fourth century free market mechanisms would seem to have operated, but in the late fourth-early fifth centuries some form of 'military contract' would appear to have taken over supply. Functional variations between different types of site have been identified with rural sites, turrets and Signal Stations sharing a major emphasis on the jar as the basic ceramic form and more complex settlement types having more diversified functional groups. The distribution of finewares also seems to be concentrated on more complex settlement types. It is apparent that there are consistent differences between the East Yorkshire region and the rest of the study area which may well reflect differences extending back into the Iron Age. Similarly there seem to be indications of some 'de-Romanisation' in late Roman assemblages but this does not develop in the fifth century, when nearly all the strands of evidence of Romano-British material culture disappear very rapidly. / Science and Engineering Research Council