Page:American Journal of Sociology Volume 15.djvu/128

 114 THE AMERICAN JOURNAL OF SOCIOLOGY

trial Revolution. The remarkable, though unfinished essay on "Ricardo" did not form a part of the history course. Three popu- lar addresses respectively on "Wages and Natural Law," "Industry and Democracy," and "Are Radicals Socialists?" are examples of his remarkable power of combining scientific statement with popu- lar address. The addresses cited were delivered to audiences of working-men and employers in important English manufacturing towns with an ardent faith in the advantage of an impartial dis- cussion of questions affecting the relations of capitalists and work- ing-men before audiences composed of members of both classes. The volume may be urged upon the attention of students chiefly for its ethical and inspirational value.

Isaac A. Loos

State University of Iowa

Die statistischen Mittelwerte. Eine methodologische Unter- suchung. Von Dr. Franz Zizek. Leipzig: Danker und Humblot, 1908. 8vo, pp. x-f-444. M. 9.

In the winter semester of 1903-4, the late von Inama-Stemegg assigned Dr. Zizek a report for his seminar in Vienna upon sta- tistical averages. From this report the present substantial volume has grown in the course of four years. i

The writer's interest is strictly confined to the "general meth- odological" aspects of averages. He does not concern himself with the purposes of statistical inquiry, with the collection of data, with the construction of diagrams, with the interpretation of tables, or with the recapitulation of results, except in so far as these matters casually affect the striking of averages. By temperament he seems to be well fitted to deal logically with this abstract subject, and his linguistic equipment is ample. But he frankly admits that an inadequate mathematical training prevents him from applying the methods of "the mathematical statisticians," or from criticizing these methods from the standpoint of their inventors.

The discussion opens with an interesting attempt to classify statistical series with reference to the problem of averages. First come series of observations upon distinct but similar units; for example the ages of individuals. Various kinds of averages are appropriate here — arithmetic means, medians, or modes. The sec- ond group embraces series whose members express the size of