creasingly alike, at least up to the time of leaving home. His general conclusions were, first, that older twins did not show greater similarity then younger twins; and, second, that brothers and sisters only a few years apart and therefore living under similar environmental conditions showed a much smaller degree of similarity than twins, the resemblance being less than half that of twins.

Professor Starch ? made an interesting experiment to ascertain the extent to which ability is the result of training. He tested 18 pairs of brothers and sisters in the University of Wisconsin for two groups of characters, those which would be affected and those which would be unaffected by school training. He found that the correlations for the former group were not appreciably larger than those for the latter group; the average of the former being .42 and of the latter 38. If these similarities had been due to training rather than to original ability, then the tests for the first group of characters would have shown a much greater degree of similarity than those of the second group.

Another experiment by Thorndike is significant as showing the effect of similar training on different individuals. Two groups of adults were tested for initial ability in one character, that of proficiency in addition. They were then put under the same course of training; and, at the end of the period, the tests showed that the two groups were farther apart than they were in the beginning. In other words the superior group had profited more by the training than had the inferior group.3 Thorndike's opinion is very positive. He says: * “Of the score or more of important studies of the causes of individual differences wh have been made since Francis Galton led the way, I do not find one that lends any support to the doctrine of human initial equality, total or approximate. On the contrary, everyone of them gives evidence that if the thousand babies born this week in New York City were given equal opportunity they would

1

still differ in much the same way and to much the same extent as they will in fact differ."

Starch has confirmed this last experiment of Thorndike's on the effect of training. He tested eight persons before and after practice in multiplication, and found at the end of the period of training that the superior individuals had made more progress from the training than had the inferior individuals.

Contributions of Statisticians. The statisticians have attempted to solve the problem of the relative importance of nature and nurture by comparisons between masses of individuals. Several thousands of school children have been measured and tested for various physical and mental characters, and the results compared, first, with the same traits in parents, or in brothers and sisters, to ascertain the strength of the factor of inheritance; and second, with various conditions of the environment, to ascertain the probable influence of the environmental factor. Although this method contains possibilities of error and is subject to abuse, nevertheless, if the material is gathered with care by experts, and if statistical fallacies are avoided, it should yield significant results. It had previously been ascertained that the correlation between parent and offspring for measurable physical traits was about .5, o representing no relationship and i perfect identity. One of the first studies undertaken was on defective vision.” And the results gave the strength of inheritance for eyesight as .49, practically the same as for the other physical characters. To ascertain the influence of the environment, four different categories were studied, "number of people to a room,” “good economic conditions,” “good physical condition of parents," and "good moral condition of parents.” In none of these categories was any perceptible environmental effect found either for boys or for girls. In fact the coefficients were more frequently negative than positive. Nor was it found that school environment was detrimental to eyesight.

The conclusion to be drawn from this study of defective vision is of course limited. It is merely that defective vision is a product of inheritance rather than of environmental conditions. It does not prove that characters less specifically physical may not be determined by environment. However, this original investigation has more recently been supplemented by other studies, and the number of correlations between individual characters and environmental conditions has been greatly extended. Investigations concerning intelligence of children showed small positive correlations with good economic conditions; that for boys being .10 and that for girls .16. Condition of glands, keenness of hearing, and diseases of the ear usually showed slight correlations with the various categories representing environmental conditions, though some of the results were negative. The results of a large number of studies of this sort were combined, and it was found that the average of the “nature" influence was .51 and that of "nurture" was .03.

A later study on the health and the mortality of infants is of special interest because environmental conditions were investigated in greater detail; and the attempt was made to ascertain the kind of environmental conditions which were most influential. For this purpose they were separated into three groups.

1. Physical factors, including dampness of house, lighting of house, overcrowding, ventilation, type of house, number of rooms, nature of sanitation, and rent.

2. Indirect parental environment, including father's occupation, regularity of work, wage of father, employment of mother, income of family, insufficiency of food.

3. Direct parental factors, including habits of father, habits of mother, cleanliness of home, use of ventilation.

A difficulty in these comparisons was early encountered when it was found that parents who were inferior physically and mentally tended to drift to the poorest tenements; therefore a correction had to be made, in so far as it was possible, to separate the influences of the physical conditions themselves from the inherited character of the parents. After corrections were made, the correlations of health of infants and of mor

1 E. M. Elderton, Relative Strength of Nature and Nurture.

tality of infants with “physical environment” and with “indirect parental environment” were almost negligible, the coefficients for mortality being even smaller than those for health. Even the uncorrected coefficients for infant mortality were very low, only about .10, though the uncorrected correlations between

health and physical factors ” in some cases rose to .3 or more. It was only in the third division of the environment, the “ direct parental factors ” that the corrected correlations were high enough to be significant. Here the average for all the factors was .154, and in one district the correlation between health of infants and cleanliness of the house rose to .33. These results give much less weight to the effect of mere physical conditions of houses than might be anticipated, but the greater importance ascribed to “direct parental factors ” is not surprising. It was to be expected that better results would be attained for children in an inferior house with intelligent parents than in a better house with parents so ignorant or careless that they failed to take advantage of the superior conditions. The conclusion which Pearson draws from the statistical data is that “it is five or ten times as profitable for a child to be born of parents of sound physique and of brisk orderly mentality, as for a child to be born and nurtured in a good physical environment.” In so far as the environment is an influence the most efficacious reforms apparently would lie in the direction of educating parents in matters of ventilation, food, and cleanliness. Moreover, it is probable that "habits of parents used in this study are themselves partly the result of inheritance, so that complete separation of the two factors would be impossible.

This method of statistical comparison has been criticised on the ground that, obtaining as it does the full force of the hereditary factor, it compares it with one factor only in the environment instead of with the total influence of the environment. It is probably true that total influence of the environment should be used in a comparison between environmental and hereditary influences on a character. But the total influence of several factors is not obtained, as Pearson points out, by adding the coefficients of correlation of each factor. The combined in

as

1

fluence is rather the resultant, so to speak, of all the factors; and such a combination would raise the measure of their influence to a small extent only, as Pearson has shown. A more pertinent criticism would be that the really decisive factor in the environment had not been properly selected or properly measured. For example, if one is trying to ascertain the relative importance of heredity and of environment on growth, he can compare the same character in parents and offspring to ascertain the hereditary influence; but there is no such character as stature in the environment, nor is there any influence which obviously determines growth. The investigator might obtain negative results for several possible influences and overlook, or be unable to measure, the most potent influence of all. Such a danger will be overcome with more extended and more careful investigations; and even now the criticism has little point if applied to recent detailed data compiled by the Eugenics Laboratory.

Arguments Emphasizing the Importance of the Environment. The large mass of material supporting the contention that the environment is paramount consists for the most part merely of reiterated assertions concerning its significance; and when the material does include statistical evidence it makes no attempt to separate the influence of the environment from that of heredity, and hence the studies can usually be interpreted in favor of either factor. Inasmuch as inferior parents tend to segregate in the worst environments, it is of little significance to show that inferior children come from bad environments. The relative contributions of heredity and of environment to their inferiority still remains a matter of doubt.

Biologists have made a considerable number of experiments, chiefly on the lower scale of animal life, showing the abnormalities which may be produced by violent changes in the environment or by unnatural environments. But, while there is undoubtedly a lesson in these experiments, they hardly cover the problem of the sociologist, for he is interested in the effects of the differences in the actual environments in which men live.

1 See J. Arthur Thomson, The Biological Theory of Nurture. Eugenics Review, 8: 50.