This article explores how diagrams are performed in actual scientific practice by examining two studies of the structure of periodic markets in rural China before and during the World War II. These are Ching-Kun Yang’s pioneering study of systematic field observations of Chinese periodic markets in the 1930s, and the US anthropologist William G. Skinner’s model, which accounts for the social and economic structure of rural China. In both cases, the empirical studies involved in uncovering the Chinese market structure are closely connected to Western theories of geographical research: Yang’s research is influenced by the Chicago School of Sociology, led by Robert E. Park and Ernest Burgess; Skinner’s theoretical research applies the central place theory of Walter Christaller and August Lösch of the German Location School. The author argues that the social scientists’ practices demonstrate how diagrams have a specific epistemic virtue for guiding the idealization and representation, both as a process and end product, in formulating explanations for the distribution of locations. Moreover, whereas the diagrams are represented as pure geometrical shapes, they not only are the outcome of idealization but also play an important role in guiding the process of evidentiation.