The nature of these data lent itself to cartographic portrayal and the mapping of data extracted from historical sources was a major element in most studies. Since 1970 computer-assisted methods of data handling have been introduced, and late nineteenth century census materials have become available. Such objective documentary sources have been complemented by increased use of qualitative sources (e. g. personal correspondence, literature and art) stimulated in part by North American work on the process and perception of colonisation and by the humanistic perspective in human geography. The general trend in research over the period 1950-85, of an increased utilisation of nineteenth and twentieth century sources and a relative decline in the use of medieval and earlier materials, resulted in a shift of emphasis from studies of rural society and landscape towards analysis of urban and industrial society. The active search for appropriate theoretical frameworks for this new focus was in contrast to the reduction in importance attached to map-based research. It is suggested that the problems associated with the construction of historical maps and atlases is an area worthy of renewed investigation. Particular difficulties such as the absence of fixed spatial units and the lack of standardised units of measurement (e.g. cartloads!) underline the fundamental influence source materials exert on analytical results. The introduction of computer techniques in historical geography has proceeded slowly and has largely been confined to the application of standard statistical packages to relatively small data sets. The full potential of computer technology to store, manipulate and analyse large data sets (such as the Domesday Book) has yet to be realised. It is concluded that historical geography's inherent conservativism with respect to data sources and methodology may be challenged by external pressures (technological change, restricted research funding, and fewer young researchers)
tended that this dependence on primary data is not sufficient to account for the lack of geographical involvement in Dark Age analyses. Several additional reasons are suggested to explain why the historical geographer's expertise in medieval studies has not been extended into the earlier period. These include the move by archaeologists away from investigations of single sites and single periods towards a 'total landscape' (geographical) perspective. Concurrently within geography the contemporary emphasis on relevance in research has acted to reduce the supply of future historical geographers, with the Dark Ages, already a minority specialisation, suffering disproportionately. Despite such difficulties, however, historical geographers have an important role to play in Dark Age studies. The complexity of reconstructing the life styles of the period means that the task is out with the capability of a single discipline. A multi-disciplinary collaborative approach is required in which the historical geographer can make a valuable contribution to the study of the 'total landscape' by bringing his own expertise to complement the excavation skills of the archaeologist. What is required is for a mutuallybeneficial symbiosis to be effected between geography and archaeology .