Non-related men differ among themselves by 5·9. Non-related women differ among themselves by 6·0. Related men differ among themselves by 4·1. Related women differ among themselves by 3·8. Relatives, and especially related women, have therefore, on the average, resemblance in reaction-type. This fact means that the psychological adaptation of relatives differs but slightly. An investigation into the various relationships gave the following:-- The mean difference of the husband and wife amounts to 4·7. The mean deviation of this mean is, however, 3·7, a very high figure, which signifies that the mean figure 4·7 is composed of very heterogeneous figures; there are married couples in whom the reaction type is very close and others in whom it is very slight. On the whole, however, father and son, mother and daughter stand remarkably close. The difference between father and son amounts to 3·1. The difference between mother and daughter amounts to 3·0. With the exception of a few cases of married couples (where the difference fell to 1·4) these are the lowest differences. In Fürst's work there was a case where the difference between the forty-five year old mother and her sixteen year old daughter was only 0·5. But it was just in this case that the mother and daughter differed from the father's type by 11·8. The father is a coarse, stupid man, an alcoholic; the mother goes in for Christian Science. This corresponds with the fact that mother and daughter exhibit an extreme word-predicate type,[155] which is, in my experience, important semeiotically for the diagnosis of insufficiency in the sexual object. The word-predicate type transparently applies an excessive amount of emotion externally and displays emotions with the unconscious, but nevertheless obvious, endeavour to awaken echoing emotions in the experimenter. This view closely corresponds with the fact that in Fürst's material the number of word-predicates increases with the age of the subjects experimented upon.