Modern geography is often said to begin with two important early nineteenth-century German scholars, von Humboldt and K. Ritter. These two men saw eye to eye in many things and agreed in pouring scorn on their predecessors because they dealt with geographical information in such a haphazard and unsystematic fashion. The reaction against ‘classical’ geography took many forms: some essentially a development from it, like the writings of Alfred Hettner; some in opposition to it, like J. Brunhes’ ideas; some developed along new lines without close reference to it. Statistical techniques are likely to be attractive to geography because they help with one of the subject’s most intractable methodological problems of the last three-quarters of a century since the decay of the ‘classical’ school, the recurring worry about the best way to accommodate in a single discipline a physical and a social side.