The present article introduces a new statistic that can be of value in the analysis of the geographic association between the members of one population (e.g., schizophrenics) and the members of another (e.g., suspected etiological factors). An application of the statistic is made to the geographic distribution of suicides in a large city. (9 refs)