2   Life and Letters of Francis Galton

simply and wholly caused by A, nor indeed by C, D, E and F as well ! It was really possible to go on increasing the number of contributory causes, until they might involve all the factors of the universe. The physicist was clearly picking out a few of the more important causes of A, and wisely concentrating on those. But no two physical experiments would-even if our instruments of measurement, men and machines, were perfect-ever lead to absolutely the same numerical result, because we could not include all the vast range of minor contributory causes. The physicist's method of describing phenomena was seen to be only fitting when a high degree of correlation existed. In other words he was assuming for his physical needs a purely theoretical limit-that of perfect correlation. Henceforward the philosophical view of the universe was to be that of a correlated system of variates, approaching but by no means reaching perfect correlation, i.e. absolute causality, even in the group of phenomena termed physical. Biological phenomena in their numerous phases, economic and social, were seen to be only differentiated from the physical by the intensity of their correlations. The idea Galton placed before himself was to represent by -a single numerical quantity the degree of relationship, or of partial causality, between the different variables of our ever-changing universe. How far he was successful forms the subject-matter of this chapter.

I have said that Galton came to this fundamental conception from two aspects. The first problem was that of inheritance. To take an illustration A character in the Father does not determine absolutely the like character in the Son ; it is only one out of many contributory factors. The character is only a partial expression of the Father's germ-plasm; so it is with the Son's character-it is not at all a full expression of his germ-plasm. Again, the Son is not a product only of his Father's germ-plasm, but of his Mother's also, and those of both parents in their turn are products of innumerable ancestral stirps leading us back through long eons of evolution. Nor is the somatic or bodily character of the Son a product only of heredity, it is the integration of a number of factors acting throughout his prenatal and postnatal growths. From the physicist's standpoint of causation there was no way at all to attack this problem, the causes were too indefinite and elusive to be individually grasped and measured. They could only be dealt with one at a time-the measure of the resemblance of offspring to parent, a partial causation, led Galton to the idea of correlation.

The second problem which impressed itself on Galton's mind was that of correlation in the narrow biological sense. The word itself appears to have originated with Cuvier who denoted by it an association between two organs or characters of a family-thus the occurrence of a split hoof with a particular form of tooth, so that from the discovery of one organ a prediction could be made as to the nature of others. It has been said that Cuvier's conception did not involve causation*. I do not know that any correlationist of to-day would assert that the knowledge of the length of the femur, which would enable him to closely predict the length of the humerus, is an assertion of

* See C. Herbst, Handworterbuch der Naturwiesensehaften, Bd. In, S. 621, Jena, 1913.