Statistical Investigations 357

C. GALTON'S ANTHROPOMETRIC LABORATORIES

The above series of schedules will show how fertile were Galton's _plans for collecting statistical data during the decade 1874-84. It was, however, only in the course of this schedule experience that he learnt how reluctant most people are to fill up a schedule. As a result of this experience Galton changed his method of action. Failing the establishment of school anthropometric laboratories Galton determined to set one up at his own cost, and catch the world when on its leisurely and inquisitive peregrinations. He called into existence the first Anthropometric Laboratory at the-'International Health Exhibition in London, 1884. On the closing of this exhibition the laboratory was removed to the Science Museum, South Kensington, and the total number of visitors measured before it was closed was well over 9000. These included both sexes and all ages from five to eighty years. This splendid material, which is only at the present time being fully reduced and utilised', together with Galton's "Family Records" embracing between three and four hundred families, some 150 'stirps,' provided him at last with the material he had so long sought. The discussion of this material furnished Galton with occupation for at least ten years ; and the need for novel statistical methods, which its problems demanded, led him to the correlational calculus, the fons et origo of that far-reaching ramification-the modern mathematical theory of statistics. One quakes to think of what might have happened had Galton not obtained through that first anthropometric laboratory and his family records the data he needed ! The latter led him at once to the quantitative measure of heredity-the correlation of kinsmen for any faculty-and the former showed him that the same problems repeat themselves in all statistical material, and that. the conception of correlation is not peculiar to heredity, but embraces all recordable qualities which without being causally linked together yet vary more or less stringently one with the other. From that conception arose anew view of the universe, both organic and inorganic, which provides all branches of science with a novum organum, far wider-reaching in its effects than that of Bacon, and as characteristic of the last quarter of the nineteenth century as the fluxional calculus was of that of the seventeenth. I have sought in vain for any forerunner of Galton in this matter', and feel convinced that he was the first to grasp not only the need of measuring associated variations, but the first to provide any real measure of them. Galton wrote to Darwin on December 24, 1869 that the appearance of the Origin of Species had formed a real crisis in his life and freed him from his old superstition as if he had been roused from a nightmare (see Vol. i, Plate II). For some of us Galton's new calculus acted in precisely the same manner; it enabled us to reach real knowledge-"to submit phenomena to measurement and number"-in many branches of inquiry where

1 See for example Koga and Morant, "On the Degree of Association between Reaction Times in the case of Different Senses," Biometrika, Vol. xv, pp. 346-72,1923.

2 See a paper by the present writer entitled " Notes on the History of Correlation," Biometrika, Vol. xiii, pp. 25-45,1920.