146 Life and Letters of Francis Galton
collected the data on which an answer as to the value of the hereditary factor could really be given.
That Galton was somewhat moved by the attitude of De Candolle to his own work is proved not only by the letters cited above, but by an interesting paper he contributed to the Fortnightly Review entitled: "On the causes which operate to create Scientific Men'." Galton begins by referring to his difficult paper "Blood Relationship" of the previous year which we shall consider later. In that paper he distinguishes clearly, if in now unusual language, between somatic and genetic characters, and in the Fortnightly he refers to the "paradoxical conclusion "-as it certainly was in 1872-that the child must not be looked upon as directly descended from his own parents. The bearing of this statement lies in the explanation it affords of the reason why children differ frequently in mental character from their parents'-an observation which had been raised as an argument against Galton's theory of the inheritance of ability. Ability in an individual marks as a rule ability in the ancestry, not necessarily in the immediate parents. Galton then turns to
0°a volume written by M. de Candolle in which my name is frequently referred to and used as a foil to set off his own conclusions. The author maintains that minute intellectual peculiarities do not go by descent, and that I have overstated the influence of heredity, since social causes, which he analyses in a most instructive manner, are much more important. This may or may not be the case but I am anxious to point out that the author contradicts himself, and that expressions continually escape from his pen at variance with his general conclusions. Thus he allows (p. 195) that in the production of men of the highest scientific rank the influence of race is superior to all others; that (p. 268) there is a yet greater difference between families of the same race than between the races themselves; and that (p. 326) since most, and probably all, mental qualities are connected with structure, and as the latter is certainly inherited, the former must be so as well. Consequently I.propose to consider M. de Candolle as having been my ally against his will, notwithstanding all he may have said to the contrary. The most valuable part of his investigation is this : What are the social conditions most likely to produce scientific investigations, irrespective of natural ability, and a fortiori irrespective of theories of heredity?
This is necessarily a one-sided inquiry But it admits of being complete in itself, because it is based on statistics which afford well-known means of disentangling the effect of one out of many groups of contemporaneous influences. The author, however, continually trespasses on hereditary questions, without, as it appears to me, any adequate basis of fact, since he has collected next to nothing about the relatives of the people upon whom all his statistics are founded.
The book is also deficient in method but it is full of original and suggestive ideas." (pp. 346-7.)
As an example of deficiency of method, Galton cites De Candolle's statistical treatment : -
"The author's tables of the scientific productiveness per million of different nations at
' Fortnightly Review, Vol. xut, New Series, pp. 345-510. No copy of this exists among Galton's collected papers and no reference to it in any of the manuscript or printed bibliographies.
' Even De Candolle (p. 282) admits that "Une aptitude naturelle est toujours probablement heritee, puisque les parents sont la cause qui a precede et determine 1'existence de 1'individu. Les exceptions s'expliquent par la diversite des parents, leur etat momentane lors de la conception " to which latter source of divergence Galton in his much annotated copy puts the note "Stuff!!" I cite this to indicate how strongly he held that the environment could not immediately affect the gametic characters, and how fully Galton rejected factors like 'maternal impressions' and `momentary states at conception' as influencing the mental character of offspring.