FAMILY MENTAL-TEST ABILITIES

271

mark; (a) may be vouchsafed a few additional words as to details: the marital coefficient lies at .44; the usual depression of the parentals below the fraternals appears, the average for the former being .35 and for the latter .42. The verbal tests show somewhat higher correlations than the non-verbals, reflecting (probably) the effect of common environment on the individuals concerned. In the individual tests, for the most. part, the sex differences are negligible, but there are a few striking exceptions, such as the superiority of the male fraternal to the female in the verbals, vice versa in the non-verbals, and the superiority of both to the mixed (brother-sister).

The growth-and-decline curves are new and instructive; they have been shown to parallel closely a considerable number of hypotheses as to mental maturity, and a smaller number of physiological findings. It was suggested that the variations in the shape of the curves from a more or less general type might prove informative with respect to the nature of different abilities and their rate and period of acquisition. The standard deviation curves were noted to rise and fall after somewhat the same fashion as the mean curves, but their interpretation is problematical.

The study has also impressed upon the writer the comparative futility of attacking this exceedingly complex problem (if results of permanent value are desired) with anything less than the entire array of philosophical, mathematical, biological, and psychological knowledge now available. Such, in fact, is its extreme intricacy, and such the degree of exclusive specialization which has up to the present time come to pass in psychology and allied fields, that it would seem that investigators interested therein would do well, besides familiarizing themselves with the intent of other disciplines, to join forces with specialists in them in some unified and organized endeavor to make a contribution sound from every possible point of view.

V

HISTORY OF THE PROBLEM

The general problem of mental inheritance is of course of great antiquity. The oldest reference of the modern period1859-available to the writer recites a curious medley of fantastic speculations on this vitally interesting topic, covering the last 2500 years; and in most cases recites them with naïve dogmatism as facts patent to any careful observer. The rise of biology, and particularly the increasing significance of the special biological problem of evolution, immensely fostered this perennial interest; especially, of course, the Lamarckian doctrine of the inheritance of acquired characters, with its ethical implications, impressed a stamp on popular thought that still, a hundred years later, constitutes a serious obstacle to intelligent action grounded on the newer discoveries. We have, for example, the numerous writings of Mr. Caspar Redfield, evidently a sincere and indefatigable worker, whose dialectic moves in this fashion: (1) acquired mental characteristics are inherited, (2) acquisition of mental abilities is cumulative with age, (3) ergo, mental improvement of the race is to be secured by breeding from older parents. And this contention is "proved" by the selection of suitable cases from history.

A more solid foundation was laid in detailed investigations into the problem of physical inheritance; here of course the beginning was made in the inclusion by Darwin of heredity in the triad (heredity, variation, natural selection) of causal factors in evolution; and a few years later by the now celebrated work of Gregor Mendel, Abbot of Brünn. Mendel was temporarily forgotten; but the stream of English thought flowed on in Francis Galton's Natural Inheritance, in his analysis of the relationships of eminent Englishmen, and in Darwin's later Descent of Man. From the first sprang also the

calculus of variations, which was to grow under the hand of Galton's pupil Karl Pearson to such primary significance as an instrument in the new investigations.

Meanwhile German and continental biology had bloomed with the rise of the Empire, and had kept closely behind English thought; and in 1900 Mendel was rediscovered by deVries and his colleagues. Investigation was profoundly stimulated in all quarters. In 1903 what was probably the fairest opportunity yet for the opening of a road leading to discoveries of the deepest significance was lost, when Pearson devoted the twelfth of his Mathematical Contributions to the Theory of Evolution to an analysis of "alternative" inheritance-and found that, with the restrictions he had imposed, the theoretical coefficients were significantly lower than those which had already been obtained. Pearson seems, from this point, to have turned definitely away from the "Mendelian" theory, and the two complementary lines of attack, having come so near unification, drifted apart again. Both, however, did work of the utmost importance; the researches of the Galton Laboratory during the first decade of the twentieth century are epochmaking, though too little known even yet; during this time the Mendelians were, so to speak, consolidating the newly won ground, testing theories and experimenting, discovering the important principles modifying the original Law of Segregation. Their great work started with the second decade-the initiation by T. H. Morgan, at Columbia, of the Drosophila investigations. In the meantime several monumental studies by the original Galtonian genealogical methods had been given to the world; Brimhall's study of the eminent relatives of Cattell's American Men of Science is a representative example in this country. The work of Binet and his followers, joining with that of the American investigators and standardizers of school procedure, has yielded numerous instruments of research and studies based thereupon; of these studies one of the most important is that on the Stanford group of gifted children, though it is potentially rather than primarily a research

in inheritance; the (yet incomplete) Burks study of fosterchildren in California also gives promise of results of the highest importance. Merriman (Stanford, 1922) showed that identical twins are more nearly alike mentally than are ordinary fraternal twins. The Eugenics Record Office under C. B. Davenport uses methods which are a combination of the Galtonian genealogical research with those of Mendelian interpretation. The investigations of Thorndike and Merriman on twins are researches in mental inheritance conducted exclusively by means of the newer standardized tests, in different stages of development. Those of Peters, Starch, Schuster and Elderton, etc., on "inheritance" of school achievement, depend either on the genealogical method or the more strictly "educational" tests.

In all this, of course, there is a noticeable absence of unified and harmonious coördination of all available knowledge in a concerted attack on the unknown; one is reminded of the two wings of an army moving forward to the assault in mutual ignorance and suspicion of each other's movements. It is gratifying to note, therefore, that notwithstanding such regrettable incidents as the polemics of the Pearson-Jaederholm study of 1913 (a bitter attack on the Mendelians, founded on an apparently complete ignorance of the principle of cumulative factors) the two branches of thought which had come so near unification in 1903 (and, less significantly perhaps, in a few sporadic articles in the interim, e.g., those of Snow and Brownlee) were tentatively fused in a theoretical development (On the Correlation between Relatives on the Supposition of Mendelian Inheritance) by R. A. Fisher in 1918. It is unnecessary to discuss this contribution extensively; it will suffice to say that it is an analysis of the statistical consequences of a more completely generalized conception than the earlier one of Pearson, and that the theoretical coefficients therein arrived at possess properties that square with the observed facts-at least, with respect to physical characteristics. The promise of such a development is of course very great; and that such

籼

promise is not without foundation is further indicated by an additional note (The Dominance Ratio, 1922) by Dr. Fisher, announcing, together with a more extended inquiry into the conditions affecting the selection and survival of genes, that certain of the simpler effects of his earlier analysis have been independently confirmed. The present writer may perhaps be pardoned a statement here of his conviction that no contribution of real significance can be made to the study of heredity in the near future which does not deal in an exhaustive and competent way with the results brought to light by these two analyses and any further work resting thereupon.