Early Anthropological Researches   127

larising twins or relatives of twins, and `snowballed' by asking them for the addresses of other twins, which he remarks led to a continually widening circle of correspondence. Finally Galton obtained information concerning 94 sets of twins (see his statement in second paper). He considers their resemblances in the case of 35 twin-sets-each of like sex-in which there was detailed evidence of close similarity. He finds that this likenessmental, physical and pathological-is maintained even when life has carried the twins into different environments ; they appear to have the same illnesses at the same times. The answers showed that in the bulk of cases the resemblance of body and mind continued unaltered up to old age and under very different conditions of life; in other cases dissimilarity was attributed wholly to some form of illness or accident which had befallen one twin and not the other. Galton then turns to the 20 sets of twins he had. detailed accounts of unlikeness between. He has not a single case in which his correspondents speak of originally dissimilar characters having become assimilated through identity of nurture.

"The impression that all this evidence leaves on the mind is one of some wonder whether nurture can do anything at all beyond giving instruction and professional training. It emphatically corroborates and goes far beyond the conclusions to which we had already been driven

by the cases of similarity. In these the causes of divergence began to act about the period of adult life, when the characters had become somewhat fixed; but here the causes conducive to assimilation began to act from the earliest moment of the existence of the twins, when the

disposition was most pliant, and they were continuous until the period of adult life." (p. 404.)

And then follows the passage cited on pA of our first volume.

There is another passage also which is of great suggestiveness and may be cited here

"Much stress is laid on the persistence of moral impressions made in childhood, and the conclusion is drawn, that the effects of early teaching generally must be important in a corresponding degree. I acknowledge the fact, but doubt the deduction. The child is usually taught

by its parents, and their teachings are of an exceptional character for the following reason. There is commonly a strong resemblance, owing to inheritance, between the dispositions of the child and its parents. They are able to understand the ways of one another more intimately than is possible to persons not of the same blood, and the child instinctively assimilates the habits and ways of thought of its parents. Its disposition is educated by them, in the true sense of the word; that is to say, it is evoked earlier than it would otherwise have been. On these grounds I ascribe the persistence of habits that date from the early periods of home education, to the peculiarities of the instructors, rather than to the period when the instruction was given. The marks left on the memory by the instructions of a foster-mother are soon spunged clean away."

(p. 405.)

Consider, says Galton, the history of the cuckoo, which is reared exclusively by foster-parents!-Neither its note, nor its habits, nor its sympathies are influenced by those of its foster-parents. Galton concludes generally that with reasonable care in the collection of our data, we may ignore the many small differences in nurture which characterise individual cases.

The reader of Galton's first paper may possibly hold that he ought to have given more of his data, but it must be remembered that his paper was written originally as a popular article for Fraser's Magazine, and Galton was not yet a practised statistician. His material is worthy of a fresh analysis,