Abstract
Roughly two decades after the first issue of the American Political Science Review, sampling theory emerged to encourage learning about the many from the few. Once social scientists had worked out the selection procedures to deal with large and far-flung human populations, the stage was set for the scientific study of citizen behavior in national elections, as initially demonstrated by George Gallup in predicting the 1936 presidential race. This development unleashed a growing torrent of literature, which by now is simply overwhelming. The many high points in this literature could scarcely be listed, much less abstracted, in the brief compass allowed this essay. So I must invoke some severe selection principles in order to proceed.