We have seen that others which increase during the whole period of growth, and are therefore subject to the continued effect of the new environment, are modified even among individuals who arrive here during their childhood. From these facts we must conclude that the fundamental traits of the mind, which are closely correlated with the physical condition of the body and whose development continues over many years after physical growth has ceased, are the more subject to far-reaching changes. It is true that this is a conclusion by inference; but if we have succeeded in proving changes in the form of the body, the burden of proof will rest on those who, notwithstanding those changes, continue to claim the absolute permanence of other forms and functions of the body.

PROBLEM OF HEREDITY.

In the course of our investigation it has been necessary to inquire into certain problems which have no immediate connection with the change of type of the descendants of immigrants, but which are of great importance for our knowledge of the amalgamation of different types in America. The most important of these is the problem of heredity.

Two theories of heredity are being held. In accordance with one, the children show a tendency to revert to a type intermediate between the types of the two parents a mid-parental type-or, in cases of changes of types, to another type dependent upon the mid-parental type. In other words, the characteristics of the parents are blended in the children. According to the other theory, the laws of heredity act rather in such a way that, in regard to certain traits, either the father's or the mother's type, or the type of a more remote ancestor, is reproduced, and that certain parental traits may be dominant over others. In a generalized way we may say that by dominance is meant the tendency of one particular trait-either the father's or the mother's to appear with greater frequency in the children than the corresponding but different trait of the other parent. It does not necessarily follow that all the traits of the same type are dominant, but dominant traits may be present in both parents, some in one, some in the other.

In a mixed population, like that of America, in which the frequencies of mixtures increase with the increasing social amalgamation of the descendants of immigrants, these questions are of prime importance, and it seemed desirable to obtain as much information as possible on this point.

An inquiry into the values of the cephalic index has shown clearly that the type of heredity in intermarriages in the same race is that of alternating heredity. Children do not form a blend between their parents, but revert either to one type or to the other. The method by means of which this problem has been investigated is based on the consideration that in case children show a tendency to revert to a type intermediate between that of father and mother, the variability of all the children in each family must remain the same, no matter what the difference between father and mother may be. On the other hand, if there is a reversion to the parental types, we must expect that the variability of the children in each family will increase with the difference in type between the two parents. The actual calculation of these data requires a reduction of the results

according to the number of children to each family. After this reduction was made and the cephalic index of the mother reduced to corresponding male values, the following results were obtained:

TABLE 17.-Difference in cephalic index between father and mother, and corresponding variability of children.

No evidence has been obtained showing that one type is dominant, but all show equal degrees of correlation between father and child and between mother and child and for various values of the cephalic index.

The exact character of the alternating inheritance can not be established from the data at our disposal. Although a large number of families have been investigated, they are all families consisting of father and mother belonging to the same type, and consequently cases of great differences between the parents are rare; and these are the ones which bring out the facts of alternating inheritance most clearly. It seems from the discussion of the available data that in the type of heredity investigated—that is, of the cephalic index among members of the same type of man-the facts agree best with the theory which assumes the absence of dominant traits in the type, an equal frequency of reversion to the parental types, and a lesser reversion to ancestral types; in short, an almost typical alternating type of inheritance.

PHENOMENA OF GROWTH.

It has also been found necessary to investigate the phenomena of growth. The older investigations of Bowditch, Roberts, Peckham, Porter, and myself, not to mention many later ones, have shown that there is a period of marked acceleration of growth during the period of adolescence. Our inquiries have demonstrated that a similar period of acceleration occurs in the growth of the head, most markedly in the growth of the length of the head, and that the period is synchronous in both sexes with the rapid increase in the bulk of the body. The occurrence of this period of acceleration of growth during the period of adolescence is of interest because the periods of rapid growth of different parts of the body are not by any means the same throughout. (Table 18, figure 22.)

The material for this investigation was obtained from the tables published by Röse, Ranke, Boas, and by use of the tables of the Immigration Commission.

a C. Röse, Beiträge zur europäischen Rassenkunde, Archiv für Rassen- und Gesellschafts-Biologie, Vol. II, 1905; Vol. III, 1906.

bOtto Ranke, Beiträge zur Frage des kindlichen Wachstums, Archiv für Anthropologie, N. S. Vol. III, p. 161 et seq.

Franz Boas and Clark Wissler, Statistics of Growth (Report of the U. S. Commissioner of Education, 1904, pp. 26, 27).

82401°-VOL 2-11-36

FIG. 22.-Rate of annual growth of length of head, width of head, and cephalic index.

a, Length of head; b, width of head; c, cephalic index.

Scale, a, b, 1 square-0.5 mm.; c, 1 square-0.5 unit

This diagram shows that both length of head and width of head have first a decreasing rate of growth, then a more rapid growth during adolescence, followed by a gradual decrease in rate of growth. The maximum rate at the period of adolescence seems to be, for girls, at about 12 years; for boys, between 15 and 16

years.

The cephalic index develops in a peculiar way. It is about constant during the early years of life, but during the period of adolescence it shows a rapid decrease in both sexes.

COLOR OF HAIR.

Special attention has been given to the determination of the color of hair, and a method has been devised for its numerical designation. It has been shown that among the lighter-haired types the color of hair of women is lighter than that of men, and that the increase of pigmentation during the period of growth is very considerable.

EFFECT OF SIZE OF FAMILY.

Among the influences upon the physical development of children that have been studied, the size of the family deserves especial mention on account of the tendency of American-born families to decrease in size. Children of small families in Toronto, Ontario, are taller than those of families of larger size. The same phenomenon exists in Oakland, Cal., where the average size of the family is much smaller than in Toronto. About the same proportions prevail among children of families of one, two, three and more children, as in Toronto. The differences obtained from the measurements in both cities, weighted according to the number of cases, are given in the table and figure following:

TABLE 19.-Stature of children of families of different sizes: Toronto, Ontario, and Oakland, California.

Number of children in family.

Excess of stature

over norm....

Weight..

+0.20

+0.21 +0.10 +0.05 +0.00 -0.05 -0.06 -0.14 -0.11 -0.12-0. 23 -0.05 -0.29 1,057 2,189 3,006 3,119 2,626 2,001 1,253 865 475 307 136 34

FIG. 23.-Comparison of the stature of children in families of different sizes.

Scale, 1 square=0.02 σ.

In this diagram the average stature of children in families of all sizes has been assumed as a norm. The number of children per family is indicated on the horizontal line; the excess or defect in stature by the distances between the broken line and the norm. It appears that children in families that have one child only have a considerable excess in stature; that the stature of children in families of from two to four sinks to the norm, while the stature of children in larger families is considerably below the normal. (The figures in Table 19 do not quite agree with those given on the diagram, on account of corrections made in the table after the diagram had been completed.)

The numerical values express the excess in stature in multiples of the standard variability. In order to reduce them to centimeters, each value ought to be multiplied by about 6. It appears from these data that the physical development of children, as measured by stature, is the better the smaller the family. It is generally assumed that the increase in stature among the groups that are better situated economically is due to better nutrition than is found among the poor. The general economic advance is also believed to account for the general increase in stature in Europe. I am inclined to lay much greater stress upon the decrease in the frequency of diseases of childhood, which have a retarding influence upon growth that can not be made up after the detrimental causes have been overcome, and it may well be that lesser frequency of diseases brings about the better development of children in small families, although the one-child families might be expected to form an exception from this point of view.