MENTAL QUALITIES

21

of inquiry. Indeed in all cases where it may be assumed that any human quality is controlled by a large number of Mendelian factors, then it can be proved by mathematical analysis that the ratios between the average measurements of groups of individuals related to each other in different ways must closely coincide with what in fact these ratios are actually found to be. In this way statistical inquiries afford strong corroborative evidence of the truth of the Mendelian theory of inheritance.1

Another most important line of inquiry initiated by Galton is that in connexion with the inheritance of mental qualities; his conclusions indicating that these mental qualities are passed on from generation to generation in accordance with the same laws which are found to hold good as regards the inheritance of physical qualities. He showed by inquiries concerning the relatives and ancestry of judges, of men of science, and of other eminent persons, how markedly talent tends to run in families. He proved not only that eminent men have as a rule eminent kinsfolk, but also that the more nearly is any group of persons related to a man of genius, the greater is the proportion of talented persons which that group generally contains. Much work has been done since Galton's day in this field of inquiry, all of which tends to confirm our belief in the natural inheritance of ability. Mental defect or feeblemindedness has also been proved to be eminently hereditary, but this is possibly more akin to the inheritance of some physical malformation than to that of definite mental attributes. Such qualities as vivacity, conscientiousness, temper, ability, etc., have, however, been included within the scope of these inquiries; whilst the inheritance of the mental attributes of animals has also been examined. Lastly, the conformation of the brain has been found to be hereditary, a fact not to be neglected since the mind and the brain are unquestionably in some way connected with each other. Nothing in all these researches tends to cast doubts on the broad conclusion that our efforts to improve the innate qualities of future 1 See R. A. Fisher, "Correlation between Relatives," Trans. R. Soc. Edin., LII, Pt. II, 399, Oct. 1918.

generations must be based on the same principles whether we are considering mental or physical qualities:

What has here been said is obviously insufficient to prove the truth of the conclusions above stated. As regards any department of science, it will become more and more apparent to the statesmen, social workers, ordinary citizens, and even to scientific students in other fields, that they are on the horns of a dilemma when deciding what conclusions to accept as valid. To form an independent judgment may require many months of continuous study, and either this labour must be undertaken, or the judgment of experts must be accepted on authority; for any position half-way between these extremes is beset with dangers. No doubt it takes some time even to ascertain not only who are the leading experts in any field of science, but also in what respects their beliefs are sufficiently unanimous in order to justify us in accepting them as valid when wishing to build reforms on reasonably safe foundations; and all that has here been attempted has been to state some of the conclusions which may thus be regarded as being authoritative. No one has a right to reject these scientific conclusions unless he can show that they are not widely accepted by experts or unless he himself has devoted some considerable amount of time to independent research.

But what is the bearing of all this knowledge concerning the laws of heredity on the eugenic problem? This will appear more clearly later on, though here it should be stated that by far the most important effect of a study of these problems, whether it be lengthy and independent or brief and based on authority, is to be found in the general impressions thus stamped on the mind. The power or inevitability of the laws of natural inheritance; how they bring within their jurisdiction all forms and all functions of organic life; how much is now known, even though much remains to be discovered; and how criminal it would be to neglect to utilize all this hard-won knowledge; these are impressions which cannot be effaced from the minds of those who have given serious thought to the biological problems of heredity.

The foregoing conclusions will doubtless not escape criticism, and to certain vulnerable points it will be necessary to return in a later chapter. Success in life was taken by Galton as a measure of talent, and especially those who have not read his works will be inclined to argue that the fact that talent' of this kind runs in families may be entirely due to family interest being able to secure good appointments for relatives, or to the advantages of social position. It may be suggested more generally that the likenesses between fathers and sons is mainly due to the similarity of the surroundings to which they have been exposed; and that until such environmental effects have been somehow eliminated, these statistical results cannot be accepted as indications of the influence of natural heredity. Then as to Mendelism, it was implied in the foregoing analogy that the factorsthe bricks of our houses-are passed on from generation to generation quite unaltered. But if this be the case, how are we to account for evolution? Again assuming a common ancestral descent for man and certain lower organisms, and also that the genes never change, it would follow that both man and certain existing jelly-like organisms are all founded on the same Mendelian factors. Is this conceivable? Here all that need be said is that these criticisms have not been overlooked.

SUMMARY

The aim of this chapter has been to give a very general account of the laws of natural inheritance, a field of discovery in which Mendel and Galton were notable pioneers. The method by which it is now believed, as the result of Mendel's work, that qualities are passed on from generation to generation was illustrated by an analogy, mainly with the object of showing that, as regards any individual, the Mendelian factors are derived from many ancestors but not from all, and of explaining how it comes about that members of the same family, though never quite alike, generally present marked likenesses. The reappearance of ancestral traits was also accounted for. The process actually visible when reproduction takes place,

including the ejection of half the chromosomes, serves as an independent confirmation of Mendelism. As to Galton's investigations, they were seen to be largely statistical, and to be of great value in enabling forecasts to be made with considerable certainty concerning the average qualities of offspring, the qualities of their ancestry being known. Galton indicated that looking only to any selected group of individuals there always exists at every generation a tendency to regress towards the mean, a tendency which is balanced by another influence acting in a contrary direction. He was also the first to give proofs that mental and physical qualities are inherited in accordance with the same laws; a conclusion confirmed since his day in several ways by many other observers. Certain controversial objections to these conclusions will be dealt with at a later stage.

NOTE. The reasons why this reversion to the mean takes place are now thoroughly understood. Supposing we were to pick out a number of exceptionally sunburnt Englishmen, we should doubtless in large measure be selecting men who had been exposed to the sun to an exceptional extent. Their children would normally not be exposed to the sun to the same exceptional extent. They would on the whole be somewhat lighter in hue, and here we should have an example of a reversion to the mean; a reversion which normally takes place in regard to every quality affected by environment. Another and a more important reason why offspring revert to the mean on the average is afforded by Mendelian dominance. Even the simplest example of reversion of this type is almost impossible to elucidate without the use of technical language. Take the case of an animal with some quality dependent on a single factor including dominant and recessive genes. With a species in a stationary condition, there will remain generation after generation the same proportion of double or homozygous dominants, of double or homozygous recessives, and of heterozygous individuals each originating from a dominant and a recessive gene. It would only be in the homozygous recessive organisms that the recessive character would be visible. At each generation some of these homozygous recessives, by mating with other forms, would give rise to heterozygous forms; and for this reason there would always be a tendency for the number of homozygous recessives to diminish. On the other hand, when the heterozygous forms mated together, a number of homozygous recessives would in consequence appear in the next generation; and here would be a tendency for the homozygous recessives to increase in every generation. These two tendencies would normally balance each other. Next let us imagine that the recessive quality was harmful, and that in a certain generation a number of homozygous recessives were eliminated by natural selection. The result of the disappearance of the organisms showing the recessive quality would obviously be an improvement in the average quality of the group.

REGRESSION TO THE MEAN

25

But this selection would upset the balance of the two tendencies mentioned above. As the result of this elimination, there would be fewer homozygous recessives, and therefore a lessened tendency for a corresponding decrease in these forms in each generation; whilst, since there would be no diminution in the heterozygous forms, there would be no lessening of the corresponding tendency for the homozygous recessive to increase. On balance the homozygous recessives would, therefore, increase somewhat; the full improvement in the quality of the group due to natural selection would not be maintained; or, in other words, when natural selection is in operation there is in each generation a partial reversion to the mean. It could be shown in like manner that whenever there is dominance, in whatever direction selection takes place, and however many factors may be involved, reversion to the mean will normally take place. Thus dominance and environment afford the two explanations why on the average the qualities of the offspring are nearer to the mean than were the mean qualities of their parents, the Mendelian theory of inheritance forming a necessary part of the explanation.