Psychological Investigations 271
new, if now familiar ideas, but it was a lecture of suggestion, and accordingly the reader must not expect to find in it statistics of actual measurements of character. It serves to explain, however, the links in Galton's own mind between his work on Heredity, his paper on Twins and his study of Free-will. At the very time Galton was writing this lecture he was collecting data by aid of his Family Records (see our Chapter XIII) on the distribution of one phase of character, namely, Temper in English Families. A First Report on his results was published in the Fortnightly Review, July, 1887'. The paper from more than one standpoint is slightly disappointing, and as Galton himself remarks he had to set to work on rough materials with rude tools (p. 29). The criticisms that one may raise are of the following kinds. The descriptions of temper are all verbal, and although many epithets are used, Galton in the main classifies into `Good Temper' and `Bad Temper.' His `Good Temper' contains not only the `forbearing' and `self-controlled' but the `submissive,' `timid' and `yielding.' His `Bad Temper' contains not only the `quick tempered,' but the `bickering' and the `sullen.' My own investigations seem to suggest a fundamental difference in `Good_ Temper' between the Self-controlled and the Weak class, and the Sullen cannot profitably be put in the same category with the Choleric. Galton does indeed make a five-group classification, namely: (1) mild; (2) docile; (3) fretful; (4) violent; (5) masterful. The distinction, however, between (1) and (2) is not that of self-controlled and weak good temper, and it is not clear whether such a marked class as the sullen has been put into (3), (4) or (5). Another defect of Galton's material was the large proportion of cases, over 50'/.,in which no record of temper was given at all. He calls these neutral and says that approximately
Good Tempered : Neutral Tempered : Bad Tempered :: 1 : 2 : 1,
and he finds in the approximate equality of the Good and Bad Tempered, and their total being equal to the Neutral Tempered, definite evidence of the correctness of the records in this respect. I fail to be convinced by Galton's arguments, for it seems to me that they would have equal application to any classification into alternate categories, e.g. criminal and non-criminal, with a neutral class for those of whom nothing was known, or nothing recorded. On the basis of his classification, omitting the 50 '/o of `neutrals,' Galton deduces that there is no selective mating in human marriage with regard to temper', but he concludes that there is emphatic testimony to the heredity of temper. His method of establishing the latter conclusion is somewhat arbitrary and somewhat elementary, but it has undoubtedly been confirmed by later work. He rather weakens his position, however, by introducing a caveat that he does not propose to deal with temper as an unchangeable characteristic. It is difficult to grasp how under such conditions it is possible to assert that temper is
"nevertheless as hereditary as any other quality." (p. 30.)
' "Good and Bad Temper in English Families," Vol. xran, N.S. pp. 21-30. A wrong year and locus are assigned to this paper in Galton's Memories, p. 328. 2 This is not confirmed by more recent researches.