Correspondence with Alphonse de Candolle 151
pre-statistical and early statistical days were essentially those of Laplace's . `esprit juste.'
Although Galton deals only with 180 selected names, he considers that the British Isles contain between the ages of 50 and 65 almost 300 men of this first class scientific status, and taking the male population between the same ages, he concludes that their frequency is about.l in 10,000. He then proceeds to consider whether scientific distinction is more frequent than this among the relatives of scientific men. Of the above ratio, however, little use is made, because it includes men who lack the same education and opportunity. What Galton actually does is to take two groups of relatives of the scientific men, namely (1) grandparents and uncles, about 660, and (2) brothers and male cousins, about 1450, and inquire the number of distinguished individuals among these; and he finds their number much greater than in the case of men with like education and opportunity in the community at large, the latter ratio being derived from school and university data. The method, though rough, is probably adequate for Galton's purpose. It should be observed, however, that his environment being taken as approximately constant, Galton is really attemptingto measure a `partial'association,which is probablyless intensethan the total association'. However both the pedigrees provided and the statistics suffice to show that men of marked scientific ability come from able families.
Galton next proceeds to analyse the qualities on which his correspondents considered their success to have depended. In each case he points out how often these qualities were present in the parents, or relatives of the scientists. The chief qualities are (1) Energy much above the average-both physical and mental; (2) Health-" it is positively startling," Galton remarks, "to observe in these returns the strongly hereditary character of good and indifferent constitutions"; (3) Perseverance-also frequently reported in the parents; (4) Practical business habits-fully a half of those who possess these accredit one or both parents with the same faculty; (5) Good Memory-"heredity abundantly illustrated," about 30 cases of especial good memory, and 13 of poor memory; (6) Independence of Character-50 correspondents possess it in excess-strong evidence of its presence in the families of these scientists; (7) Mechanical Aptitude-found to exist largely in chemists, geologists, biologists and statisticians as well as in physicists and engineers. Galton then turns to religion and religious bias, having previously pointed out that the clergy are badly represented among men of science (pp. 25-26) or among their parents (p. 24). On this point it seems to me that such a question: "Has the religion taught in your youth had any deterrent effect on the freedom of your researches?" can scarcely be accurately answered by the subject-himself. If he has cast off the religion of his youth, he may believe that he thinks freely; on the other hand if he is still a devout believer he would be unlikely to admit that his religious views hampered his scientific research. The scientific authors of the Bridgewater Treatises no doubt would have said their science was not the poorer for their religious bias, but we, who
' For example, if we endeavoured to measure the resemblance between brothers by taking sons of 6 foot fathers only, we should reach a value 40 7, too small.