Toward an Expanded Central-Place Model

Abstract

W A ,rALTER CHRISTALLER'S system of central places was posited as an explanation of the spatial distribution, the number, and the size of a specific kind of urban place within a large area.I The notions that each central good has a range over space and that the population sizes of central places, when arrayed, group around certain modal values forming typical size classes are of fundamental importance to his construct.2 However, Christaller's model, although a great pioneering effort, is imperfect. One source of imperfection is the fact that some of the basic parts of his system are not separable. For example, he first attempted to demonstrate the existence of unique types of central places when those places were arrayed according to their populations, and then, employing the towns in the size class that had the smallest typical population, he determined a typical minimum distance between central places. Given this information, plus the notion that each central good has a spatial range, he deduced typical magnitudes for the distances between the places lying within the various size classes; each typical size class had associated with it a typical distance value, and as population sizes increased, so did distances.3 Obviously, the distances that he determined were significantly affected by the size types already established. The fact that the magnitudes of the typical distances are dependent on the determination of typical population-size classes is a serious weakness. Various mathematical and statistical models have been presented that also deal with the number and population size of cities4 found in a large area.5