June 5^ 1914] 



SCIENCE 



829 



summary of progress in the study of variation, 

 but giving some elementary explanations of 

 methods. 



Of more recent works those of the greatest 

 value have appeared in the English language. 

 It vcould have been too much to expect that 

 the influence of the newer work would extend 

 as far as Madrid and affect the " Tratado Ele- 

 mental de Estadistica " of Menguez y Vicente 

 (1907) but something better might have been 

 expected of France and Germany. How back- 

 ward they are can be seen from such essays as 

 Zizek's " Soziologie und Statistik," Miinchen 

 und Leipzig, 1911, or those in the " Fest- 

 schrift" for Georg von Mayr, "Die Statistik 

 in Deutschland nach ihrem hevitigen Stand," 

 Miinchen und Berlin, 1911. Such works as E. 

 Faure's " Elements de Statistique " and 

 Maciejewski's " ISTouveau Fondements de la 

 Theorie de la Statistique " (1911) have, in 

 spite of their attractive titles, no practical 

 value to the biologist — or, as far as the present 

 reviewer can see, to any one else. Nor can 

 anything be said in favor of Al. Kauffmann's 

 " Theorie und Methoden der Statistik," just 

 published (Tiibingen, 1913). 



Forscher's recent book* seems to be a serious 

 attempt, apparently done in almost complete 

 ignorance or with all but a total disregard of 

 the fundamental calculus of the English 

 school, to obtain a better theoretical (mathe- 

 matical) basis for statistical formulae. Of the 

 soundness of the mathematical work, I am not 

 able to judge ; nor does it particularly concern 

 us here, for practically — that is, as a tool for 

 the man grappling with practical statistical 

 problems — the work has little immediate value. 



Turn now to introductory works which may 

 be of value to the beginner. 



Many biologists and others are very deeply 

 indebted to the sections on Heredity, Selection 

 and Evolution in the second edition of Pear- 

 son's " Grammar of Science " for their first 

 knowledge of the new methods. The second 



* Forsoher, H., "Die Statisehe Methods als Selb- 

 standige Wissensehaft. Eine Einfiiliruiig in 

 deren Fimdamente und Grundziige, " Leipzig, 

 1913. 



edition is out of print, but a third is being 

 issued. 



" The Primer of Statistics " (London, A. & 

 C. Black, 1909) prepared by the Eldertons at 

 the suggestion of Sir Francis Galton is for the 

 general public rather than for one who hopes 

 to prepare for investigation. King's little 

 book° although containing something of the 

 newer methods is for the social rather than for 

 the biological student and is far too elementary 

 for those who wish to do research work. Elder- 

 ton's book' not only embodies the actuaries' 

 viewpoint, but presupposes the actuaries' train- 

 ing. Thus as a first book it is quite beyond 

 the depth of the average biologist, but if he is 

 working seriously in statistics it is most help- 

 ful on his table. 



That Eugene Davenport's book^ on breed- 

 ing — with all its points concerning which the 

 doctors in mathematics and the doctors in 

 biology will disagree — is a pretty good intro- 

 duction for a certain class of readers is per- 

 haps sufficiently established by the influence 

 which it has evidently exerted in our agricul- 

 tural institutions. One recognizes the scissors 

 and paste nature of the volume, but this char- 

 acteristic it shares with practically all the 

 other introductions to statistical methods: the 

 compilation is only a little more obvious and 

 the compiler distinctly more honest in the 

 acknowledgment of original sources than is 

 sometimes the case. 



Thorndike' several years ago prepared an 

 introduction for the use of psychologists which 

 has recently (1913) been issued in a miich- 

 amplified second edition. The work is more 

 elementary and far more verbose than the book 

 by Brown.* For this reason, Thomdike's text 

 5 King, W. I., ' ' Elements of Statistical 

 Method," New York, Maomillan, 1912. Cf. Sci- 

 ence, N. S., Vol. 36, p. 519, 1912. 



<5 Eiderton, W. P., ' ' Frequency Curves and Cor- 

 relation, " London, C. & E. Layton, 1906. 



7 Davenport, E., ' ' Principles of Breeding, ' ' 

 with an appendix on Statistical Methods by H. L. 

 Eietz, 1907. 



8 Thorndike, E. L., "An Introduction to the 

 Theory of Mental and Social Measurements," 

 New York, 1904. 



Brown, W., ' ' The Essentials of Mental Meaa- 



