CTLF Corpus de textes linguistiques fondamentaux • IMPRIMER • RETOUR ÉCRAN
CTLF - Menu général - Textes

Boas, Franz. Race, Language and Culture – T03


Report on an Anthropometric Investigation
of the Population of the
United States 11

Characteristics of the Population of the United States

The White population of the United States differs from that of
Europe not so much in character as in the mode of assemblage of
its component elements. The important theoretical and practical
problems that arise in a study of the biological characteristics of our
population relate largely to the effects of the recent rapid migrations
of the diverse types of Europeans. The problem is further complicated
by the presence of a large Negro population, of small remnants of
Indian aborigines, and by a slight influx of Asiatics.

It would be an error to assume that the intermingling of different
European types is a unique historical phenomenon which has never
occurred before. On the contrary, all European nationalities are highly
complex in origin. Even those most secluded and receiving the least
amount of foreign blood at the present time have in past times been
under entirely different conditions. An excellent example of this
kind is presented by Spain. The Iberians are the earliest substratum of
population with which we are acquainted. The coast population was
undoubtedly affected by a certain amount of intermixture with
Phenician and Greek colonists. There followed a number of migrations
of Keltic tribes from northwestern Europe and a thorough
colonization of the peninsula by Rome. The Teutonic tribes which
invaded Spain came in part from the regions of the Black Sea. Later
28on we can trace waves of migration from northern Africa, which
attained their greatest importance during the time of the Moorish
empires. With the development of medieval conditions and the expulsion
of the Moors and the Jews, the population of Spain became stable
and there was no further disturbance due to important migrations. It
is therefore evident that the present population of Spain contains elements
derived from practically all parts of Europe and from northern

Similar conditions may be observed in Great Britain, where there is
also clear evidence of a large number of waves of migration. In prehistoric
times we find a long-headed type, quite different in appearance
and in customs from a later round-headed type. With the beginning
of historic times we observe first Roman colonization, then waves of
migration entering Great Britain from all parts of the North Sea, from
Scandinavia and northern Germany, and, finally, the influx of the
Normans. With this event extended migration ceased and the population
of the island was gradually welded into the modern English.

Migrations of this kind may be recognized even in very early times.
After sweeping over the older population of Greece, north European
types established themselves in the Balkan Peninsula and on the Aegean
Islands during the so-called Doric migration, which occurred a thousand
years before our era. Later on the movements of the Finnish
ancestors of the Bulgarians and the migrations of the south Slavic
peoples added to the intermixture of types in the eastern European

It might seem that a few countries in Europe were not so much
exposed to intermixture as those previously mentioned, and it is particularly
assumed that Sweden and Norway represent a very homogeneous
population. Still, we may recognize here also a considerable differentiation
of local types. An investigation of the districts nearest to
Finland shows very clearly an approach to the Finnish type which may
be due to intermixture. In southern Norway is encountered a strongly
aberrant type whose origin cannot be historically determined. In the
northern area the Lapps present a foreign element. In later times immigrations
were not by any means rare. Thus the development of the
mining industry brought in a great many Walloons; and the nobility,
at least, is a composite of descendants of natives from many parts of
Europe. Historical evidence shows that the central parts of Europe
29over which migrations have swept periodically were, even more than
the outlying districts, exposed to intermixture of different types.

Intermixture in Europe was largely confined to antiquity, although in
some parts it continued into the Middle Ages, whereas the intermingling
of different local types in the United States is recent. Owing
to the social conditions in ancient Europe amalgamation of distinct
elements may have been rather slow. Notwithstanding the relatively
small numbers of migrating individuals, it may have taken several
generations for the intrusive and native populations to become merged.
In the United States, owing to the absence of hereditary social classes,
the amalgamation is on the whole more rapid and involves larger
numbers of individuals than the intermixture which took place in
earlier periods in the Old World.

The impression that the population of European countries is comparatively
speaking “pure” in descent is founded on its stability. In
northern and central Europe this condition developed after individual
hereditary landholding was substituted for the earlier forms of agricultural
life, and with the attachment of the serf to the soil which he inhabited.
These conditions prevailed in the Mediterranean area even
in antiquity, but in the northern parts of Europe they did not develop
until the Middle Ages, when the more or less tribal organization of
the people gave way to feudal states. During the period when the Keltic
and Teutonic tribes moved readily from place to place a vast amount
of mixture occurred in all parts of Europe. Later on, when families
became settled, those parts of the populations which were proprietors
of the soil, or otherwise attached to the soil, became stationary, and
consequently intermixture between distant parts of the continent became
much less frequent than in previous times. On the other hand,
the mutual permeation of neighboring communities probably became
much more thorough.

These conditions of stability continued until by the development of
cities diverse elements were brought together in the same community.
This process became important with the growth of modern industrialism
and with the concomitant growth of urban populations that were
drawn together from large areas. Investigations made in different parts
of Europe, particularly in Italy 12 and in Baden, 23 show differences in
type between city populations and those of the open country. These
30may in part be explained by the strong intermixture of types drawn
from a wide area which assemble and intermarry in the city. Observations
of the population of Paris 14 indicate the same kind of intermixture
of north European and central European types.

The settlement of the unoccupied districts of the United States has
brought about an intermixture of types similar to that occurring in
modern city populations, because settlers from different parts of
Europe may dwell in close proximity in newly opened countries. Although
in many cases we find a strong cohesion of farmers who come
from the same European country, there is also a great deal of scattering.

It should, therefore, be understood that the problems presented by
the population of the United States do not differ materially from the
analogous European problems. The differences are due to the larger
numbers of individuals involved in the whole process, in its rapidity,
in its extension over rural communities, and in the forms of cohesion
between members of the same group which are dependent upon the
mode of settlement of the country. The process resembles earlier
European mixtures in so far as many diverse European types are involved.
In modern Europe only European types enter into the mixture,
but a number of races morphologically removed from the White
race enter into certain phases of the problem in America. Even this
aspect of the problem was probably present in antiquity when slaves
of foreign races formed a considerable part of the population.

The long continued stability of European populations which set in
with the beginning of the Middle Ages and continued, at least in rural
districts, until very recent times, has brought about a large amount of
inbreeding in every limited district. In default of detailed statistical
information in relation to the development of populations it is impossible
to give exact data, but a cursory investigation shows that inbreeding
of this type must have occurred for a very long time. The theoretical
number of ancestors of every living individual proceeds by multiplication
by two from generation to generation back, so that ten generations
(or approximately 300 or 350 years) ago every single individual would
have had 1,024 ancestors. Therefore, about 600 or 700 years ago
there would be more than 1,000,000 ancestors for each individual.
Considering the stability of population, and the fact that brothers and
sisters have the same ancestors, such an increase in the number is, of
31course, entirely impossible, and it necessarily follows that a very large
number of individuals in the ancestral series must be identical, which
means that there must have been a large amount of inbreeding.

The “loss of ancestors” becomes the greater the further back we go
in the ancestry and the more stable the population. It is obvious that
particularly in the landholding group of families which remains from
generation to generation in the same place, there must have been much
inbreeding. Statistical information is available only for a few village
communities and for the high nobility of Europe. The genealogies of
all these families demonstrate that the decrease in the number of ancestors
is very considerable. The calculations for the high nobility of
Europe 15 show that in the sixth ancestral generation there are only 41
ancestors instead of 64; in the twelfth generation, only 533 instead of
4,094. These numbers seem to be quite similar to those found in the
stable village communities of Europe. Owing to this intermixture and
to the similarity of descent of the families constituting the population,
each family represents fairly adequately the whole population, or as
we might express it, the whole population is homogeneous, in so far as
all the families have the same kind of descent. On the other hand, in
a population that results from recent migration and in which individuals
from the most diverse parts of the world come together, a single family
will not be representative of the whole population, because entirely
different ancestral lines will be present in the various families. Therefore
the population will be heterogeneous in so far as the different
families belong to different lines of descent. To illustrate this point
we might assume a community consisting of Whites and Negroes in
which the Whites always intermarry among themselves, and the
Negroes among themselves. Obviously in such a population, a single
family would not be representative of the whole community, but only
of its own fraction. On the other hand, if we had a community in which
Whites and Negroes had intermarried for a long time, as is the case
among the so-called Bastards of South Africa — a people very largely
descended from Dutch and Hottentots and in which this intermingling
has continued for a long time — we have a homogeneous population
in so far as every family represents practically the same line of descent. 26
32It will therefore be seen that homogeneity is not by any means identical
with purity of race. In the case of a homogeneous population of mixed
descent we may expect, on the whole, a high degree of variability in
the family, while all the families will be more or less alike. On the
other hand, in a heterogeneous population in which each part is, comparatively
speaking, “pure,” we may expect a low variability of each
family with a high variability of the families constituting the whole
population. On account of its migratory habits the American city population
must be heterogeneous. Heterogeneous are also the immigrants
and their immediate descendants, whereas in the stationary populations
of New England villages and of the Kentucky mountains we have presumably
homogeneous groups.


For determining the characteristics of a population knowledge of the
laws of heredity is indispensable. Ordinarily the term heredity in
relation to racial 17 characteristics is used in a somewhat loose manner,
and we should distinguish clearly between the hereditary stability of a
population and the hereditary characteristics which determine the
bodily form and functions of an individual. The concept of hereditary
stability in a population can mean only that the distribution of forms
which occur in one generation will be repeated in exactly the same
way in the following generation. This is clearest in the case of a
homogeneous population as defined before. In every population varying
bodily forms of individuals will occur with characteristic frequencies.
In an undisturbed homogeneous population we must necessarily
assume that each generation will show the same characteristic distribution
of individual forms. If it did not do so there would be a disturbance
of the hereditary stability.

Conditions are quite different in a heterogeneous population like that
of the United States. Owing to intermarriages between the various
constituent types there must be a tendency toward greater homogeneity,
setting aside, of course, the influx of new immigrants. Experience shows
that no matter how rigid may be the social objection to intermarriages
between different groups, or how strong the pressure to bring about
marriages between members of the same group, they will not prevent
the gradual assimilation of the population. An instance of this kind is
33presented by the castes of India in Bengal. Notwithstanding the rigid
endogamy of castes it has been observed that the highest castes are
similar in type to the peoples of Western Asia, while the lower down
in the scale of castes we go the more this type becomes mixed with the
older substratum of the native population. 18 This can be explained only
by intermarriage between the different castes which must have occurred
notwithstanding the rigid laws forbidding it. The less the tendency
toward segregation of different groups, the more rapid will be the
approach toward homogeneity. Therefore notwithstanding the laws of
hereditary stability in individual strains, there cannot be a hereditary
stability of a heterogeneous population until homogeneity has been
attained. It may even be considered doubtful whether a disturbance
of the distribution of bodily forms may not occur as an effect of the
intermingling of two populations similar or even identical in type, but
of different ancestry, in which, therefore, a heterogeneity of ancestry
exists. 29

Thus it will be seen that the physiological laws of heredity are quite
different from the statistical expression of the effects of heredity upon
a large population. The latter depends upon both the biological laws
of heredity and the peculiar social structure of the population which is
being considered. These two aspects of heredity must be kept clearly

Unfortunately, the laws of heredity in man are not clearly known,
and it is not yet possible without overstepping the bounds of sound,
critical, scientific method to apply them to the study of the characteristics
of a population. A considerable amount of preliminary fundamental
work must be done before we can proceed to the explanation
of special complex phenomena. One fundamental point of view may
be considered as established, namely, that when a definite couple of
parents is given, the probability of occurrence of a given form among
the descendants of this couple is fixed. In man it is not easy to demonstrate
this fact because the number of children for each couple is
small. If we assume, however, an organism in which each parental
couple has an infinitely large number of offspring, the laws of heredity
may be so expressed that each form that occurs among the offspring
34has a definite probability. In man these laws can be investigated only
by combining many families in which both parents, or at least one of
the parents, has the same characteristic form, although in this case the
phenomenon is obscured by the fact that the same form in the parent
does not necessarily mean the same ancestry. 110 Observation of various
features of the body of man shows that the simple forms of Mendelian
heredity are not often applicable. It is true that in a number of cases
of pathological modifications, the validity of the simple Mendelian
formulas has been established. Even in these cases the number of observations
is not sufficient to determine whether we are dealing with exact
Mendelian ratios or with approximations. Practically all other cases
are still open to doubt. Even in the case of eye color, which has been
claimed to be subject to a simple Mendelian ratio with dominance
of brown over blue, the available figures are not quite convincing. 211
For the more complex variable measurements of the body simple
Mendelian ratios are certainly not applicable. Up to the present time
the complex laws governing the frequencies of occurrence of bodily
forms among descendants of an ancestral line are not known.

The investigation of any population must, therefore, take into consideration
the detailed study of the laws of heredity.

The Influence of Environment

In settling in the United States the immigrants have been brought
into a new environment, geographically as well as socially, and the
question arises whether the new environment exerts an influence upon
bodily form and functions. It has been customary to consider certain
features of bodily development as absolutely stable, and anthropologists
have characterized modern human types as “permanent forms” which
have lasted without variation from the beginning of our modern
geological period up to the present time. It is fairly easy to show that
in this view exaggerated importance is ascribed to the phenomena of
observed hereditary stability.

We know that the bulk of the body of an adult depends to a certain
extent upon the more or less favorable conditions under which the child
35grows up. It has been shown that malnutrition or pathological conditions
of various kinds may retard growth, and that the retardation
may be so considerable that it cannot be made up by long continued
growth. As a matter of fact, the bulk of the body at the time of birth is
so small as compared to the bulk of the body of the adult that it is easy
to understand that environmental conditions must exert a considerable
influence upon its development. Proof of this is the gradual increase
of stature during the past fifty years, until 1914, which has been demonstrated
by investigations in a number of countries in Europe, and
the difference in stature which is found in the same nationality for
people living under different economic conditions. 112

Since many proportions of the body are related to stature and bulk,
these will also undergo modifications due to environmental conditions.
The influence of environment is not so obvious in those cases in which
the bodily form is practically determined at the time of birth, or in
those in which the total growth from the time of birth until the
adult stage is very slight. It might be assumed that in all cases of
this type heredity alone determines the characteristic form of the

From a wider point of view the assumption that environment has no
influence upon the form of the body does not seem justified. It must
be understood that the question of stability or instability of the body
in relation to environmental influences has no relation to the question
of the inheritance of acquired characteristics. Even if we should adhere
most rigidly to the dogma of the impossibility of the transmission of
acquired characteristics, we must admit that a modification of the
bodily form of the individual is easily conceivable without the necessity
of assuming any modification of the germ plasm owing to individually
acquired variation. We should rather have to say that adaptability of
a definite type is one of the hereditary characteristics of the germ plasm.
The problem involved is readily understood in the case of plants which
appear in strongly modified form according to the environment in which
they grow. In many cases the amount of hairiness, the form of the
leaves, etc., are subject to the degree of moisture of the soil, and an
accurate description of the species would therefore involve a statement
that the plant has a certain degree of hairiness, dependent as a definite
36function upon the moisture of the soil, or that the leaves have a certain
form dependent upon outer circumstances. In other words, the plant
has a definite form only under a definite environment, and with changing
environment, the form changes.

We may include under the group of environmental effects also all
those variants of form and function that are dependent upon social
habits which influence the organism. An influence upon bodily form
is exerted by the habitual uses to which groups of muscles are put.
Thus the rest position of the lower jaw is different in different areas.
The English seem to hold the lower jaw a little farther forward than
the Americans. The people of the western states relax the soft palate
more than those of the North Atlantic area. The facial expression is
determined by the development of the groups of facial muscles; the
variations of certain aspects of the form of the hand and the foot are of
this kind. The functioning of organs is even more markedly dependent
upon habits, particularly upon habits firmly established during childhood.
This is illustrated by the characteristic gait of individuals and
of whole groups of people; by the involuntary movements in response
to certain stimuli; by many of the expressive movements of the body;
by habits of articulation; and by the dexterity and accuracy of movements
obtained by early training.

Since we recognize the influence of environment upon the form of
body including such features as bulk of body, or muscular forms and
the functioning of organs, it seems justifiable to define racial characteristics
as we do those of a variable plant, namely, by stating that under
definite environmental conditions the bodily form of a race and its
functioning are such as we observe, without prejudging the question in
how far modifications in form and function may result from changing
environment. The actual problem, then, would be to determine whether
and how far the traits of the body may be so influenced. We
should also bear in mind that it is perfectly conceivable that there may
be congenital modifications in forms which are nevertheless not hereditary. 113
Constitutional changes in the body of the mother may bring
about modifications in prenatal growth which to the superficial observer
might give the impression of hereditary changes. These considerations
demonstrate that it is necessary to consider this problem in any thorough
investigation of the characteristics of the American population.37


The question must be asked in how far selective agencies may determine
the movements of the population, including immigration and
emigration, the settlement of the western parts of the United States by
the inhabitants of the eastern states, and the migration from country
to city. Besides migration, the selective influences of mating, of mortality,
and of fertility have to be taken into account. Of late years
much stress has been laid upon the effect of selection upon the constitution
of a population.

The effect of selection as determined by bodily form can be investigated
to advantage only in a homogeneous population. When every
family may be considered as representative of the whole population,
and when all strata of society present the same physical characteristics,
selective forces that are based on social stratification will not influence
the selective results, because all social strata will be alike. If it
should be found that groups representing different bodily forms have
different tendencies to migrate, or different rates of mortality or fertility,
we might have an expression of the direct dependence of selection
upon bodily form.

As a matter of fact, however, homogeneous populations do not exist
anywhere in the world. A greater or less amount of heterogeneity has
always been observed, and heterogeneity in our modern civilization, at
least, is always connected with social stratification. In a heterogeneous
population like that of the United States the difficulties in the way of
determining a direct relation between selective influences and bodily
form are almost insurmountable. If, for instance, descendants of a
certain nationality are attracted to a particular area, as the Scandinavians
to the northwest, the Hungarians to the mines of Pennsylvania,
the Mexicans to the southern borderland of the United States, or the
French Canadians to the New England states and northern New York,
we must remember that each one of these social groups represents a
certain physical type and that there will be, therefore, an apparent relation
between selection and physical type which in reality is based on
social factors.

Similar observations may be made with regard to selective mating.
Since mating depends upon social contact, marriages will occur among
the groups that associate together. Wherever nationalities cluster together,
where denominational or racial considerations act as endogamic
38restrictions, there will be selective mating of similar types due to social
heterogeneity. Besides this there may be a certain amount of selection
that unites tall with tall or expresses the sexual attractiveness of other
bodily features.

Social heterogeneity exerts an influence also upon the mortality and
the fertility of different groups. The more recent immigrants are on
the whole less well-to-do than the earlier immigrants and their descendants.
We know that there is a relation between fertility and economic
well-being and we find, therefore, that the number of children of the
more recent immigrants is greater than that of the descendants of
earlier immigrants, so that, setting aside the question of mortality, there
would be a shifting in the distribution of the population in favor of later
immigrants. Since the earlier immigrants represent the northwestern
European type and the later immigrants the south and east European
types, there will appear in this case also a selection according to bodily
form, which is due not to the direct relation between physical characteristics
and fertility, but rather to the fact that the one economic group
is composed of one type, and the other economic group of another type.
In many cases the relation between descent and social stratification
is so complex that it easily escapes our notice, and for this reason we
may observe phenomena of selection apparently related to bodily form
but actually due to obscure social causes that are discovered with great
difficulty only.

On the other hand it cannot be denied that in some cases at least
there must be a direct relation between bodily form and physiological
function on the one side and selective processes on the other. It is, for
instance, quite obvious that in the settlement of the new western countries
a certain bodily and mental vigor was necessary to enable a person
to undertake the venture. It has often been pointed out, although it
has never been proven empirically, that in this way there must have
been a selection from the inhabitants of the New England villages who
migrated westward and that the emigrants represented a physically
superior type. Even though this conclusion is not based on observation
it seems highly probable. To the same group of phenomena would
belong the supposed greater susceptibility to certain forms of disease
of slightly pigmented individuals, as compared with the greater power
of resistance of brunet individuals. I am not by any means convinced
that incontrovertible proof of this assumption has been given; but if
it were true that the constitution of the blond is weakened by exposure
39to intense sunlight, there might be a selective influence of this kind when
a people move from the cloudy temperate zones to the brilliant sunlight
of more southern and more arid climes.

In considering the selective influences of environment it should be
borne in mind that the human body is so constituted that all its organs
can operate adequately under widely varying circumstances. Our lungs
are able to supply the needs of our body under the air pressure that
prevails at the level of the sea, and they operate adequately at an elevation
of 20,000 feet where the air is highly rarefied. The heart can
adjust itself to the variation in demands made upon it, either in sedentary
life at the level of the sea, or in active life in high altitudes. Our
digestive organs may adapt themselves to a purely vegetable diet or
again to a purely meat diet. Our central nervous system is also capable
of adjusting itself to the most varied conditions of life. As long, therefore,
as the conditions of environment do not exceed very elastic limits,
it is not probable that selective influences would become operative to
any very great extent, at least not in so far as they are determined solely
by the form and functioning of the organs of the body.

Racial and Individual Differences

An investigation of the bodily forms of the individuals constituting
a race, homogeneous or heterogeneous, shows that they differ considerably
among themselves in every single feature, such as pigmentation,
form of hair, size and proportions of the body, physiological reactions.
These differences are measurable and express the degree of variability
of the race. A complete presentation of the characteristics of a race
would contain a statement of the relative frequency of each particular
bodily form which occurs among the individuals constituting the race.
When comparing, from the point of view of anatomical or physiological
characteristics, the racial types of Europe which constitute the bulk of
the American population, it appears that the range of variation for the
different types is of such a character that a great many individuals belonging
to one type correspond to other individuals belonging to another
type. In other words, there are certain forms common to all populations
of Europe. To give an example: We find strongly contrasting
head forms in northern Italy and in Sardinia. Nevertheless an investigation
of the distribution of head forms in each one of these districts
shows that 27 per cent of the population may belong either to Sardinia
or to northern Italy. In other words, there is a very considerable
40amount of overlapping of bodily form between neighboring types, and
it is only when we consider races that are fundamentally different that
we find certain characteristics that do not overlap. Comparing, for
instance, the blond north European White and the dark Sudanese
Negro, there is no overlapping with regard to pigmentation, form of
hair, form of nose, form of lips, etc. If, on the other hand, we proceed
by steps from northern Europe to the Sudan, a great many intermediate
and overlapping steps between these extreme forms will be found, so
that only the extremes would really be entirely separate. While it may
be that two races are quite distinct with regard to certain features, there
are always other features with regard to which the differences are so
slight that the assignment of any one individual to either one race or
the other would be beset with doubt.

It has been customary to express the differences between racial types
by the difference between the averages of each type or between the
modes (the most frequent values) that are characteristic for each type.
It is easily shown that such a description in misleading. If we wish to
express the difference between two individuals, each of whom has
constant characteristics, we may proceed in this manner. If one individual
measures 170 cm. and another 165 cm., the difference between
them is 5 cm. If, however, a certain population has an average stature
of 170 cm., and another population an average stature of 165 cm., we
cannot say that the difference between the two is 5 cm., because if
there is a wide range of variability there will be a large number of
individuals among the taller population who have exactly the same
statures as individuals of the shorter population. To give arbitrarily
selected figures, the one may range perhaps from 150 to 190 cm., the
other from 145 to 185 cm. In this case an individual that measures
anywhere between 150 and 185 cm. might belong to either class. It
must, therefore, be clear that if we speak of differences between two
races we do not necessarily mean differences between individuals, and
these two concepts must be kept clearly apart. The bulk of our modern
literature concerning racial differences is open to misinterpretation
owing to a lack of a clear understanding of the significance of the term
“difference” as applied on the one hand to individuals and on the
other hand to races. The generalization, which is often made (to
use our previous instance), that the one population is 5 cm. shorter
than the other is often interpreted as meaning that this implies a characteristic
of all the individuals of a race, while actually a single selected
41individual of the shorter race may be much taller than a single selected
individual of the taller race. This is equally true of all those anatomical,
physiological, and psychological characteristics which exhibit overlapping
of individuals. It is also true of those that show no overlapping,
because the difference between two selected representative individuals
may vary within wide limits. If it is stated that the Whites have larger
brains than the Negroes, this does not mean that every White person
has a larger brain than any Negro, but merely that the average of the
Negro brains is lower than the average of the brains of the Whites.
With regard to many characteristics of this kind, we find that the difference
between the averages of different races is insignificant as compared
to the range of variability that occurs within each race.

An additional point should be considered in connection with this
phenomenon. Most of the anatomical characteristics of the body are
stable throughout adult life, until senile degeneration begins. On the
other hand, physiological and psychological functions are not the same
in the same individual at all times. They vary strongly with environmental
conditions and particularly with different demands made upon
the organism. The variability of physiological and psychological responses
is therefore much greater than the variability of anatomical
form, because the two former combine the variability due to the difference
in the functioning in various individuals with the variations of
response under varying conditions. When comparing racial types we
must therefore avoid expressing a difference of types simply as a difference
of averages.

Another point must be considered which may be illustrated by an
example. Let us assume that in one area the color of the hair varies
from black to dark brown with an average value on a certain definite
shade, and that in another population the color of the hair varies from
dark blond to very light blond with an average on a certain shade of
blond. In this case the two distributions will not overlap at all. On
the other hand, let us assume that we have two populations with the
same average shades of brown and of blond as before, but in the one a
variation which begins with black and extends into blond shades, and
in the other a pigmentation which begins with a very dark brown and
extends into very light blond, so that the two overlap. Obviously the
two differences will not impress us as the same, notwithstanding the
fact that the two averages remain the same. It is therefore indispensable
that in an investigation of this kind the significance of the difference
42between two populations should be clearly expressed, and that the impression
should be avoided that the difference between racial types is
identical with the difference between individuals.

Still another point deserves attention. Many writers assume that an
individual of a certain type represents the same biological type regardless
of the racial group to which he belongs. To give an example: a
round-headed person of the Tyrols is equated with a round-headed
person of southern Italy, at least in so far as the form of the head is
concerned. Even if we assume that the round-headedness of the two
individuals is of the same kind, this inference is not tenable. It is true
that by chance the two individuals may belong to the same lines of
descent, but a study of a series of homologous individuals shows that
genetically, and therefore physiologically, they are not the same notwithstanding
the sameness of the particular trait that is made the subject
of study. When we select, for instance, individuals with the same
head index of 82 in a population that has the average head index of
85, the children of the selected group will be found to have an average
head index of 84; when we select individuals with the same head index
of 82 in a population that has the average head index of about 79, the
children of the selected group will be found to have an average head
index of about 80, for the reason that there will be in each case reversions
to the average type of the population to which the selected group
belongs. In other words, the individuals which are selected from any
population must always be considered as part of this population and
cannot be studied as though they were an independent group.


One of the reasons for the special stress that is laid upon race investigations
is the fear of race degeneration. It is assumed that the intermixture
between different racial types and the rapid increase of the
poorest part of the population have a deteriorating effect upon the
nation. In the introductory remarks I have tried to show that there is
little reason to believe that racial intermixture of the kind occurring in
the United States at the present time should have a deteriorating effect.
I do not believe that it has been adequately proved that there is a
clearly marked tendency toward general degeneration among all
civilized nations. In modern society the conditions of life have become
more varied than those of former periods. While some groups live
under most favorable conditions that require active use of body and
43mind, others live in abject poverty and their activities have more than
ever before been degraded to those of machines. At the same time the
variety of human activities is much greater than it used to be. It is
therefore quite intelligible that the functional activities of each nation
must show an increased degree of differentiation, a higher degree of
variability. Even if the general average of the mental and physical
types should remain the same, there must be a larger number now than
formerly who fall below a certain given low standard, and also a larger
number who exceed a given high standard. The number of defectives
can be counted by statistics of poor relief, delinquency, and insanity,
but there is no way of determining; the increase of those individuals who
are raised above the norm of a higher standard, and they escape our
notice. It may therefore very well be that the number of defectives
increases without influencing the value of a population as a whole,
because it is merely an expression of an increased degree of variability.

Furthermore, arbitrarily selected absolute standards of value do not
retain their significance. Even if no change in the absolute standard
should be made, the degree of physical and mental energy required
under modern conditions to keep oneself above a certain minimum of
achievement is greater than it used to be. This is due to the greater
complexity of our life and to the increasing number of competing individuals.
Greater capacity is required to attain a high degree of prominence
than was needed in other periods of our history. The claim that
we have to contend against national degeneracy must, therefore, be
better substantiated than it is now.

The problem is further complicated by the advance in public hygiene
which has resulted in lowering infant mortality and has thus brought
about a change in the composition of the population, in so far as many
who would have succumbed to deleterious conditions in early years
enter into the adult population and must have an influence upon the
general distribution of vitality.

Notwithstanding the doubtful basis of many of the assertions relating
to degeneracy, the problem of eugenics is clearly before the public, and
the investigation of racial and social types cannot be separated from the
practical aims involved in the eugenic movement.

The fundamental thought underlying eugenic theory is that no environmental
influences can modify those characteristics which are determined
by hereditary nature. Nurture, it is said, cannot overcome

We should recall here what has been said before regarding the difference
between the characteristics of hereditary strains and those of races,
and that while it is true that strains differ greatly in physical and mental
vigor and in specific characteristics, it is not equally true of races as a
whole, because strains which are very much alike in all these characteristics
are found in every single race. Even if it is not possible to prove
with absolute certainty the complete identity in mental traits of selected
strains belonging to races as diverse as Europeans and Negroes, there is
not the slightest doubt that such identity prevails among the various
European types. Eugenics, therefore, cannot have any possible meaning
with regard to whole races. It can have a meaning only with regard
to strains. If the task of the eugenist were the selection of that third
of humanity representing the best strains, he would find his material
among all European and Asiatic types, and very probably among all
races of man; and all would contribute to the less valuable two-thirds.

As an objection to this point of view it is sometimes claimed that
closely allied animal types are so different in their physical make-up and
mental characteristics that members of one race can be clearly differentiated
from those of another race. It is, for instance, said that the race
horse and the heavy dray horse are so different in character that no
matter what may be done to the dray horse its descendants can never
be transformed into race horses. This is undoubtedly true, but the
parallelism between the races of dray horses and race horses on the
one hand and human races on the other is incorrect. The races of horses
are developed by careful selection, by means of which physical and
mental characteristics are fixed in each separate strain, while in human
races no such selection occurs. We have rather a racial panmixture,
which brings it about that the racial characteristics are distributed
irregularly among all the different families. As a matter of fact, dray
horses and race horses correspond to family strains, not to human races,
and the comparison is valid only in so far as race horses and dray
horses are compared to the characteristics of certain family lines, not to
human races as a whole. In Johannsen's terminology the human races
are to a much greater extent phenotypes than races of domesticated

For this reason the task of eugenics cannot be to devise means to
suppress some races and to favor the development of others. It must
rather be directed to the discovery of methods which favor the development
of the desirable strains in every race.45

This problem can be attacked only after the solution of two questions.
First of all, we have to decide what are the desirable characteristics;
and secondly, we must determine what characteristics are hereditary.
With regard to the former question, we shall all agree that physical
health is one of the fundamental qualities to be desired; but there will
always be fundamental disagreement as to what mental qualities are
considered desirable — whether an intense intellectualism and a repression
of emotionalism or a healthy development of emotional life is preferable.
Obviously, it is quite impossible to lay down a standard that
will fit every person, every place, and every time, and for this reason
the application of eugenic measures should be restricted to the development
of physical and mental health. Even if it were possible to control
human mating in such a way that strains with certain mental characteristics
could be developed, it would seem entirely unjustifiable for our
generation to impose upon future times ideals that some of us may consider
desirable. It might furthermore be questioned whether the interests
of humanity will be better served by eliminating all abnormal strains
which, as history shows, have produced a number of great men who
have contributed to the best that mankind has done, or by carrying the
burden of the unfit for the sake of the few valuable individuals that may
spring from them. These, of course, are not scientific questions, but
social and ethical problems.

For the practical development of eugenics it is indispensable to determine
what is hereditary and what is not. The ordinary method of
determining heredity is to investigate the recurrence of the same phenomenon
among a number of successive generations. If, for instance,
it can be shown that color-blindness occurs in successive generations, or
that certain malformations like polydactylism are found repeatedly in
the same family, or that multiple births are characteristic of certain
strains, we conclude that these are due to hereditary causes; and if
parents and children have the same head form or the same or similar
statures, we decide that these similarities also are due to heredity. It
must be recognized that in many of these cases alternative explanations
are conceivable. If, for instance, a family lives under certain economic
conditions which are repeated among parents and children, and if
these economic conditions have a direct influence upon the size of the
body, the similarity of stature of parents and children would be due to
environment and not to heredity. If a disease is endemic in a certain
locality and occurs among parents and children, this is not due to
46heredity but to the locality which they inhabit. In other words, wherever
the environmental conditions have a marked influence upon bodily
characteristics, and wherever these environmental conditions continue
for a number of generations, they have an effect that is apparently identical
with that of heredity. In many cases the causes are so obvious that
it is easy to exclude persistence of characteristics due to environment.
Under other conditions the determination of the causes is not so easy.

It is still more difficult to differentiate between heredity and congenital
features. For example, if a child before birth should be infected
by its mother, there might be the impression of a hereditary disease,
which, however, is actually only congenital in the sense that it is not
inherent in the structure of the germ plasm. Although the distinction
between environmental causes as previously defined and hereditary
causes is generally fairly easy, the distinction between congenital causes
and true hereditary causes is exceedingly difficult, in many cases impossible.
The long continued discussions relating to hereditary transmission
of disease are a case in point. Most of these questions cannot
be solved by statistical inquiries, but require the most careful biological
investigation. The conditions, however, are such that we must demand
in every case a clear differentiation among these three causes.

There is little doubt that in the modern eugenic movement the assumption
of hereditary transmission as a cause of defects has been exaggerated.
Although certain mental defects that occur among well-to-do
families seem to be determined by heredity, the mental defects generally
included in eugenic studies are of such a character that many of
them may readily be recognized as due to social conditions rather than
as expressing specific hereditary traits. A weakling who is economically
well situated is protected from many of the dangers that beset an individual
of similar characteristics whose economic condition is not so
favorable, and it must be admitted that criminality in families that may
be mentally weak and which are at the same time struggling for the
barest subsistence is at least as much determined by social conditions
as by heredity. Investigators of criminal families have succeeded in
showing frequencies of occurrence of criminality which are analogous
to frequencies which may be due to heredity, but they have failed to
show that these frequencies may not as well be explained either wholly
or in part by environmental conditions. We should be willing to admit
that among the poor undernourished population, which is at the same
time badly housed and suffers from other unfavorable conditions of
47life, congenital weakness may develop which lowers the resistance of
the individual against all forms of delinquency. Whether this weakness
is hereditary or congenital is, however, an entirely different question.
Experiments made with generations of underfed rats 114 suggest that a
strain of rats which has deteriorated by underfeeding can be fed up
by a careful amelioration of conditions of life, and it may well be questioned
whether delinquent strains in man may not be improved in a
similar way. Certainly the history of the criminals deported to Australia
and of their descendants is very much in favor of such a theory. In
other words, it seems very likely that the condition of our subnormal
population is not by any means solely determined by heredity, but that
careful investigations are required to discriminate between environmental,
congenital, and hereditary causes.

Formulation of Problem

From the preceding discussion, we may formulate the principal problems
that must be taken up in a study of the population of the United States.
We have to investigate first the degree of homogeneity of the population;
second, the hereditary characteristics of the existing lines; third, the influence
of environment; fourth, the influences of selection. On the basis of
the data thus collected, we have to interpret the significance of the differences
between various types, and investigate the bearing that our results
may have upon public policies.

The study of the adult population alone would not give us adequate data
to enable us to clear up the causes which determine the final development
of the body — the events which take place during the period of growth must
also be taken into consideration.

Familiarity with the bodily forms of children is necessary also from a
morphological point of view. On the whole, the development of individuals
is divergent, so that the most characteristic forms of each type are
found in the adult male. The adult female forms are not quite so divergent,
perhaps in part for the reason that the period of development of the
female is shorter than that of the male, although it must be remembered
that secondary sexual characteristics are present in childhood. The younger
the human form that we investigate, the less clearly are racial characteristics
expressed. We may, therefore, say that the most generalized forms
of a racial type will be found in the infant or, even still more clearly, in
prenatal stages, while the most highly specialized local forms will be found
in the male adult. A knowledge of the specialized forms ought to include,
therefore, a study of progressive differentiation. Particularly for the study
of the influences of environment it is indispensable that the development of
the body in childhood should be studied while the influences are still at
48work. We have to know the conditions which bring about retardation or
acceleration in the development of various parts of the body, and their
ultimate effects upon the human form. We must study other minute
changes that may perhaps not be related to retardation or acceleration, but
that may be due to a direct effect of environmental causes. In the adult
these changes have been completed and can no longer be subjected to
analysis, while in the growing child, their gradual development and unfolding
may be observed.

The same is true with regard to selection. If selection is related to
bodily form, it will probably act with particular intensity during the early
years of childhood. It might be revealed by a comparison of the surviving
and dying parts of the population of various ages.

These considerations make it quite necessary to include in the study of
the population, not only adults, but also children.

One method of approach should consist, therefore, in the study of the
growth and development of children, classified according to descent and
geographical and social environment. If it were feasible to include records
of the longevity of the individuals measured in childhood, the problem of
selection could also be attacked. In the study of adults a careful classification
according to descent and social position will be necessary.

The phenomena of homogeneity and of heredity make it necessary that
the investigation should not be confined to studies of individuals, but that
the anatomical characteristics of families should be made the subject of

A considerable amount of work has been done by many investigators,
throwing light upon a number of aspects of the problems here discussed.
The earliest and most extensive series of observations was collected in connection
with the War of the Rebellion and was published by Gould and
Baxter. 115 Their well-known statistics, which have been quoted again and
again, give data with regard to the stature of enlisted men according to
their nativity, descent, and occupation, and reveal the facts that inhabitants
of different parts of the United States differ in their physical development;
that the differences between the various European nationalities are
repeated here; but that in every single case, the members of a certain nationality
exceed in bulk of body the corresponding European series; and,
finally, that certain differences may be observed between groups of individuals
following different occupations.

The next important inquiry relating to our subject was an investigation
of school children of Boston by Henry P. Bowditch, 216 in which similar
differences appeared. Bowditch also showed that the differences between
various nationalities persisted throughout the period of growth, and that
49marked differences are found according to social stratification. Classification
of the population according to the occupation of the parents showed
a better development among the commercial and professional classes than
is found among unskilled labor. Soon after Bowditch's investigation similar
inquiries were instituted by Peckham 117 in Milwaukee, and later on in
a number of other cities — Worcester, Mass.; 218 St. Louis, Mo.; 319 Toronto,
Canada; 420 Oakland, Cal., 521 etc. On the whole, the methods pursued were
similar to those applied by Bowditch, and the results proved the occurrence
of analogous phenomena. Porter, in his investigation in St. Louis,
added to his inquiries the problem of the relative development of the children
of varying mental achievement, and demonstrated a difference in the
development of what he called precocious and dull children. Work of this
type was gradually taken up by educational institutions and the effort was
made to correlate physical development with school work, with a view to
demonstrating a practical way of assigning a child to his proper developmental

In similar investigations in Europe attention had been called to the fact
that the measurement of children of different ages and the calculation of
a growth curve on this basis does not give us adequate information with
regard to the details of the phenomena of growth, and it was pointed out
that repeated measurements of the same individual are necessary to obtain
fuller records. In spite of numerous efforts that have been made to obtain
such series, it has not been possible up to the present time to follow out
the development of the same individual from childhood to adult life, at
least not in numbers that are sufficient for a clear understanding of the
phenomena involved in this process.

A certain amount of material bearing upon stature and weight has been
collected by life insurance companies. This, however, is probably to a
great extent so uncertain that it is only of slight use for scientific investigations.
Military statistics taken in the United States since the War of the
Rebellion are not numerous and not very extensive. A certain amount of
work was done during the recent war, but the results have only now been
made accessible. The only fairly extended investigation of families that
has been undertaken in the United States was made in connection with the
50work of the Immigration Commission, during which a fairly large number
of Jewish, Bohemian, Italian, and Scotch families were studied in such a
manner that the phenomena of heredity could be considered in some detail.

We have practically no material whatever bearing upon the facts of
racial mixture. It is particularly worth remembering that there are hardly
any investigations to speak of that bear upon the physiological development
of the Negro and Mulatto population. In view of the ever-repeated
claim that the Mulatto is inferior in physical development to either the
pure Negro or to the White, and considering the large number of Mulattoes
in our population, it seems of fundamental importance that an investigation
of this kind should be made.

Although less important from a practical point of view than the Negro
problem, race mixture between Whites and Indians has received some
attention. Material collected in 1892 shows that the half-blood, so far as
fertility and stature are concerned, is superior to the full-blood Indians. 122
The observations relating to fertility were confirmed by the material collected
in the census of 1910. 223 Recently an inquiry into the characteristics
of the half-bloods of Minnesota was made by Professor Albert E. Jenks. 324
We are still lacking, however, full investigations into the anatomical and
physiological characteristics of half-bloods.

The problem of the intermixture between Negro and White and Negro
and Indian has hardly been touched at all. A few studies of Negro children
and soldiers do not contribute much to our knowledge. A systematic study
of the problem was made by Felix von Luschan in 1915, but the results of
his observations are not yet available. Another important inquiry is that by
Eugen Fischer on the Rehobother Bastards, the descendants mainly of
Dutch settlers and Hottentots in South Africa. This is the only work in
which the anthropological characteristics of the Mulattoes have been
taken up in detail. The theoretical as well as the practical importance of
the investigation of the Mulatto question can hardly be sufficiently emphasized.
On the one hand, we may hope to obtain by this means an insight
into the laws of heredity in man. On the other hand, the well-being of so
many millions of citizens of our country is involved that the most painstaking
inquiry should be demanded. This is the more urgent since many
States have regulated race intermixture by laws which are based simply
upon public prejudice without the shadow of knowledge of the underlying
biological facts — without even the knowledge of the peculiar form of
racial intermixture that characterizes the relations between Whites and
Negroes in the United States. In by far the greater number of cases the
51mother is a Negress and the father a White man. This results in an infusion
of White blood into the Negro race without affecting materially the
White race. A searching analysis of the hereditary characteristics of the
racial groups has not yet been made. It is true that the records of morbidity
suggest typical physiological differences, but considering the fact
that similar differences are found between different social groups of the
same race, it is not possible without further investigation to distinguish
definitely between the influences of heredity and of social environment. 125

I refrain from giving a detailed bibliography and review of the anthropometric
material collected in the United States in view of the very excellent
collection of titles made by Professor Bird T. Baldwin, of the Bureau
of Child Study of the University of Iowa. 226

Proposed Investigations

The first and most fundamental inquiry that has to be made relates to
a description of the various types constituting the population of the United
States. As explained before, it will not be sufficient to describe the adult
male and female forms, but it will also be necessary to determine the course
of growth and development which is characteristic of each form. In order
to carry through this inquiry it is necessary to obtain information with
regard to the forms characteristic of each moment of the period of the
development, and to determine the sequence of the characteristic developmental
stages of each type. It is not admissible to assume that the physiological
conditions which are found in a six-year-old Italian child must be
the same as those of a six-year-old Scandinavian child. Furthermore, the
individuals of each racial group will differ among themselves considerably
with regard to the time when certain stages of physical development are
reached, and it is therefore necessary to investigate fully the variability of
physiological development characteristic of each group. It must be considered
one of the most urgent aims of an investigation to determine the sequence
of events and the racial and environmental conditions that influence
them. There are indications that these problems may be found to be exceedingly
intricate. An example may illustrate this point. The development of
poor children is considerably retarded. Nevertheless, the second dentition
among these children is accelerated. This may perhaps be due to less care
given to the deciduous teeth and their earlier loss which stimulates the
appearance of the permanent teeth — or it may be due to other causes. It is,
however, an indication that the sequence of events indicating the physiological
changes in the body are subject to quite diverse causes.

The determination of all phenomena of this kind is very difficult when
the attempt is made to derive data by the so-called generalizing method,
that is to say, if we merely collect information that children of a certain
age show the stage of development in question so and so often, and if we
52try to derive the rate of development by subtracting the relative frequency
of occurrence observed in one year from the relative frequency of occurrence
in the next year. If we observe, for instance, that a certain tooth is
present in 50 per cent of the children of one age and in 70 per cent of
another set of children who are one year older, and conclude that in 20
per cent of the children the tooth in question will erupt in the course of
that year, the different composition of the annual groups and the different
numbers observed make it difficult to obtain reliable results. It is almost
indispensable that for each individual there should be noted the moment
of occurrence of the physiological change which is being studied. Material
of this type is almost non-existent.

The movability of our modern city populations causes great practical
difficulty in the organization of this work. It is not easy for an investigator
to remain in touch for a sufficiently long time with the same children,
and so many children change from one place to another that an initial
number of, let us say one hundred, who are studied when five years old, will
have dwindled down to an insignificant number at the time when the adult
stage is reached. For this reason an elaborate organization is needed to
carry through this work completely. To a greater or less extent, the work
must be pieced together of fragments. For children of school age, roughly
speaking from four or five years to fourteen years, the investigation might
be organized. For older children of high-school age, it will also ordinarily
be possible to carry through the inquiry, and in certain cases the transfer of
a subject from high school to college may also be followed up. It is, however,
obvious that the individuals who can be followed in this way are a
group selected according to economic and social conditions. Those groups
of the population which are well-to-do and which lay great stress upon the
acquisition of a good education will be represented much more fully than
other groups. The observations for different ages will therefore require a
consideration of the different composition of the series. An organization
like the Child Study Bureau of Chicago or the corresponding organizations
in Iowa City and Detroit, or the Association for Improving the Condition
of the Poor in New York, will be best able to control inquiries into these

The most difficult problem encountered in these investigations is the
differentiation between hereditary differences and those due to retardation
or acceleration. To give an example: A boy twelve years old may be tall
because his hereditary characteristics are such that he belongs to a tall
stock, or it may be that he is tall on account of an acceleration of his development.
Since, furthermore, the adult stature of the individual will depend
not only upon his hereditary characteristics, but also upon environmental
conditions that have an effect upon the acceleration or retardation of his
growth, it is difficult to determine directly how much is contributed by
hereditary and how much by environmental causes. From a practical point
of view the demand is always made that the anthropometric investigation
of the individual shall differentiate between these two causes.53

In most cases, however, it seems almost impossible to do so, except by a
very detailed investigation of the physiological conditions of the body.
Measurements are always subject to alternative explanations, as being due
either to hereditary causes or to acceleration or retardation, while physiological
changes are not so likely to be fundamentally different for different
hereditary lines. If, for instance, in a certain individual the loss of healthy
deciduous canines should be very much retarded, we should have the right
to assume that, in whole or in part, his bodily development may be influenced
by retardation. This, of course, presupposes a previous investigation
which would show that the hereditary characteristics of different strains do
not show very great differences in the time element of the loss of the first
canine, provided the environmental conditions remain the same. Here,
again, we are entirely lacking in material that would enable us to answer
this question, and it is evident that a very considerable amount of information
would have to be amassed in order to enable us to solve the problem. It
does not seem a hopeless task to determine the contributory effect of retardation
and acceleration in an individual child, but it presupposes a much
more thorough knowledge of the sequence of the developmental stages
than we now possess.

The description of racial types cannot be considered complete without
an inquiry into the homogeneity or the heterogeneity of the series. It is clear
from the remarks made on pages 32 et seq. that this problem can be solved
only by an investigation of the forms represented in fraternities, because
homogeneity can be proved only by showing that the types represented in
different families are the same. In other words, the investigation of homogeneity
must be based on an inquiry into the variations presented by different
families. The small size of the human family makes it necessary to see
to it that the proper weight is given to each fraternity in accordance with
its numerical composition. 127 This investigation must be supplemented by an
inquiry into each fraternity, the variability of which will depend upon the
more or less composite character of its ancestry.

We are thus led to a consideration of the problem of how far it is possible
to discover relative unity or multiplicity in the ancestry of a racial type.
The method to be pursued will depend entirely upon the laws of heredity
involved. In those cases in which we have some kind of Mendelian inheritance — that
is to say, a tendency of certain traits of the offspring to revert
to either parental type — we must obviously expect a higher degree of variability
in the mixed types than the one found in the pure parental types.
Attention has been called to the occurrence of such a phenomenon with
regard to the head index of Italians. 228 The short-headed north Italians are,
comparatively speaking, uniform in type, and the long-headed south Italians
are also fairly uniform in type, while in the intermediate regions in
which undoubtedly the two types have intermingled for a long period, the
54variability of the head index is very much increased. In a similar way, we
find that there is an increase of variability in Sweden in those regions in
which there is an admixture of foreign types that are more short-headed
than the Swedes. 129 It has also been shown that in those cases in which father
and mother belong to the same racial group, but in which they represent
extreme head forms, the one extremely short-headed, the other extremely
long-headed, the variability of the children is greater than in those cases in
which the parents represent nearly the same type. 230 We recognize, therefore,
that when a Mendelian reversion occurs, increased variability may
indicate composite descent.

There are, however, other cases in which the results of mixture have not
the effect of increasing variability. Statistics of half-blood Indians have
shown that the width of face, which is great in the Indian race and very
small in the White race, has an intermediate value among the half-bloods,
with a marked tendency, however, of reversion to a form that is narrower
than the face of the pure Indian and wider than the face of the pure White,
while the values for width of face which are half-way between the characteristic
values of the Indians and of the Whites are not so frequent as
the two other values previously mentioned. There is, therefore, a certain
kind of reversion in this case. Nevertheless, the total variability of the
width of the face of the half-bloods is almost the same as that of the pure
parental types. 331 If we assume in this case a pure reversion to either type,
we should find that the variability would be considerably more than that of
the parental races. It appears, therefore, that we cannot generalize with
regard to the phenomenon, and we have not the right to assume that mixture
will always be accompanied by increased variability and that slight
variability does not always indicate purity of descent.

Whenever the laws of heredity are of a still different type, the variability
may be affected in a very different manner. Thus it has been shown that
mixture between Europeans and Indians results in a stature which exceeds
that of the pure Indians, which in turn is greater than that of the pure
Whites. It is obviously quite impossible to predict in this case what the
variability of the series may be.

Another method of investigating mixed descent of a race is by means
of a study of the correlations of different measurements of the body. To
give an instance: when two peoples intermingle, one of which has very
long and very narrow heads, while the other has very short and very broad
heads, and if, furthermore, reversion to parental forms obtains, then we
55must expect that among the individuals representing this population very
broad heads are commonly associated with shortness, while very narrow
heads are associated with greater length. We should, therefore, expect of
such a population that the broader the head, the shorter it will be. In other
cases, where we have a single line of descent, the condition is reversed. The
size of the head depends upon the bulk of the body, and since in such a case
broad heads are indicative of length of body, the length of head is also
increased, and we find that a broad head is associated with greater length
of head. In the case mentioned before, the reversion of the normal correlation
is indicative of mixed descent.

Here again many variations may occur. If, for instance, we had two
races intermingled with a tendency to reversion to parental forms in which
the heads of one group are very high and at the same time narrow and
short, while those in the other group are flat and at the same time broad
and long, then the result would be that in the study of the correlation between
length and breadth of head, the correlation would appear inordinately
high, because all the shortest heads would belong to the high type
and would, therefore, also be narrow, while all the longest heads would
belong to the low type and would also therefore be broad.

A characteristic case in which heterogeneity of a series causes abnormal
correlation between physical features is found in Italy. Normally there is
practically no correlation between hair color and stature, but in Italy the
tall Alpine type has lighter hair than the short Mediterranean type. In
Piedmont where the Mediterranean type is practically absent we find the
following distribution:

tableau Stature | Hair | Red | Blond | Brown | Black | Less than 160 | 160-165 | 165-170 | 170 and more

In Sicily where the Alpine type is practically absent we find:

tableau Stature | Hair | Red | Blond | Brown | Black | Less than 160 | 160-165 | 165-170 | 170 and more

In Venice and Latium, on the other hand, where mixed types occur, we

tableau Stature | Hair | Venice | Latium | Red | Blond | Brown | Black | Less than 160 | 160-165 | 165-170 | 170 and more56

and still more clearly for the whole Kingdom:

tableau Stature | Hair | Red | Blond | Brown | Black | Less than 160 | 160-165 | 165-170 | 170 and more

On account of the greatly varying laws of heredity it is impossible to predict
which method of inquiry will lead to a satisfactory result. By some of
the means here described the problem of pure or mixed descent may be

The study of any particular type will, therefore, require a multiplicity
of investigations, the most important of which relate to the development of
the racial type from childhood on, the homogeneity of the series, and the
purity of the ancestry. For these purposes the investigations of children
and of families are indispensable and must accompany a generalized investigation
of the population as a whole.

For the study of the influence of environment the investigation of growing
children is, if anything, more important than for the investigation of
racial characteristics. After the adult stage has been reached environment
will not exert any further influence. The earlier in life the investigation can
begin, the more likely we are to obtain adequate results.

In this investigation the generalizing method of comparing local types
or types presented in social strata is of little use, because in order to establish
definitely an influence of environmental causes, we must be certain
that the hereditary composition of the populations which we study is the
same. For instance, when we compare a rural and an urban community,
there is nothing that will guarantee to us that both populations are derived
from the same ancestry. On the contrary, we may assume that the urban
population is drawn from a wider group than the rural population. In the
same way, when we compare the inhabitants of a long secluded valley and
find differences in bodily form between the people living in the lower part
and those living in the upper part, the question would arise whether the
ancestry of the two groups is the same and whether the people in the upper
regions have not been more isolated than those farther down. It is on the
whole easier to exclude obvious environmental influences in an investigation
of racial types than to exclude differences of racial descent in studies
of the influence of environment. The only way to escape from these complications
is by confining the studies strictly to a comparison between
parents and children.

It has been explained before that in a number of cases we may find
apparent hereditary traits which may be deduced from the similarity of
parents and their own children, and which nevertheless are primarily due
to environmental causes. If we should find, for instance, a low stature
among individuals who have been undernourished as children, and if the
next generation will also be undernourished, we may have an apparent
57similarity in stature which is not due primarily to heredity, but rather to
the fact that the same environmental causes act upon the parental group
and upon the group of children. In most cases these elements cannot be
eliminated unless we have the opportunity to study the same racial type in
different forms of environment.

It has been stated before that a modification of bodily form due to environment
which is observed by comparing parents and their children does
not contradict the phenomena of heredity. If we find, for instance, that the
stature of Jewish immigrants into the United States is lower than that of
their children, the hereditary stability of stature will nevertheless manifest
itself. The children of an exceptionally tall couple who exceed the average
stature of the immigrant Jew by a certain amount may be expected to show
an excess of stature which is correlated to the excess of stature of the
parents, which, however, has to be added to the increased average stature
of the children of immigrants. In short, a change in type due to environmental
influences simply means that the correlated deviations in the group
of parents and of children must be reckoned from the point which is typical
for the generation in question.

In some cases in which the environmental influences are very strong, a
generalizing method may give adequate results. Bowditch, in his investigation
of Boston children, was able to show that Irish children differ in their
development according to the economic condition of the parents, and there
is little reason to doubt the uniformity of the genetic composition of his
various Irish groups. But whenever the differences involved are slight, and
when they may be equally well explained on the basis of difference in genetic
composition, the comparison between parents and children is indispensable.
The data for the study of environmental influences must, therefore, be
based on the comparison of the bodily forms of parents and their offspring.
In this manner the doubt as to the difference in genetic composition may
be eliminated, although it is at least conceivable even in this case that there
may have been selective rather than environmental influences. It might be
said, for instance, that when some parents have children in charitable institutions
while other children stay at home, differences between the two
groups of children might not be due to environmental influences only but
also to selection. This example indicates that care must be taken to eliminate
the influences of selection even when we are dealing with family
groups in which diversity of genetic descent has been excluded.

At the present time it is unknown to what extent the influences of environment
may determine bodily form. Notwithstanding the numerous
claims of the fundamental effect of climate upon the body of man, we have
no evidence whatever that will show that pigmentation undergoes fundamental
changes under climatic conditions; that the White race would become
darker in the tropics; or that the Negroes would become lighter in
the north. Whatever statistics we have on this subject show rather a remarkable
stability of pigmentation. We have not even any definite indication
that the pigmentation of the hair undergoes changes under different
58climatic conditions, although in this case the change in color from the
period of childhood until middle life is so great that we might very well
expect environmental influences to express themselves. On the other hand,
we know that the bulk of the body is very susceptible to environmental influences,
and it is but natural that retardation or acceleration during the
period of growth will also leave its effect upon those proportions of the
body which depend upon bulk. Other changes which occur very early in
life are not so easily explained. I think the evidence showing that the form
of the head is susceptible to environmental influences is incontrovertible. I
also believe that adequate proof has been given for modifications in the
width of the face under changed conditions of life. The causes of these
changes are still entirely obscure. It may well be, as suggested by Harvey
Cushing, that chemical changes occur under new environmental conditions
and unequally influence growth in different directions. This would agree
with the changes in chemical constitution found in lower animals living in
different types of environment, if it is true that changes of this kind do
occur and modify the form of body so fundamentally that according to
the ordinary schemes of classification a people might be removed from one
group and placed in another one, then we have to consider the investigation
of the instability of the body under varying environmental conditions as
one of the most fundamental subjects to be considered in an anthropometric
study of our population.59

11 The following paper deals with purely anthropometric problems. It was intended
to show the kinds of information needed for understanding the meaning of
bodily build of individuals in relation to their descent and social environment. For
this reason the important questions relating to relative fertility as bringing about
changes in the constitution of the population and the problems involved in the hereditary
characteristics of pathological characteristics, physiological and psychological
traits determined by the genetic character of the individual were not touched upon.
The whole problem should be solved by a consideration not only of the anthropometric
traits, but also by a detailed study of heredity, of functions of the body and
of the differential constitution of the population. Since the paper was written much
valuable work has been done in this direction, particularly by the Population Association
of America. Journal of the American Statistical Association, vol. 18 (June,
1922), pp. 181-209.

21 Ridolfo Livi, Antropometria Militare (Rome, 1896), p. 87 et seq.

32 Otto Ammon, Zur Anthropologie der Badener (Jena, 1899), p. 641.

41 Franz Boas, “The Cephalic Index,” American Anthropologist, N. S., vol. I
(1899), p. 453.

51 Ottokar Lorenz, Lehrbuch der gesammten wissenschaftlichen Genealogie (Berlin,
1898), p. 289 et seq., pp. 308, 310, 311.

62 Eugen Fischer, Die Rehobother Bastards (Jena, 1913); Franz Boas, “On the
Variety of Lines of Descent Represented in a Population,” American Anthropologist
N. S., vol. 18 (1916),p. 1 et seq.

71 The terms “race” and “racial” are here used in the sense that they mean the
assembly of genetic lines represented in a population.

81 H. H. Risley and E. A. Gait, Census of India, 1901 (Calcutta, 1903, vol. 1,
pp. 489 et seq.

92 M. D. and Raymond Pearl, “On the Relation of Race Crossing to the Sex
Ratio,” Biological Bulletin, vol. 15 (1908), pp. 194 et seq.

101 Franz Boas, “On the Variety of Lines of Descent Represented in a Population,”
loc. cit.

112 Helene M. Boas, “Inheritance of Eye Color in Man,” American Journal of Physical Anthropology, vol. 2 (1919), pp. 15 et seq.

121 Rudolf Martin, Lehrbuch der Anthropologie (Jena, 1914), p. 225. Second
edition (Jena, 1928), vol. 1, p. 297.

131 Cf. infra p. 47.

141 Helen Dean King, Studies on Inbreeding (Wistar Institute, 1919).

151 B. A. Gould, Investigations in the Military and Anthropological Statistics of American Soldiers (New York, 1869).

162 H. P. Bowditch, “The Growth of Children,” 8th Annual Report, Massachusetts
Board of Health
(Boston, 1875), pp. 273-323; 10th Annual Report (1879), pp.
33-62; 21st Annual Report (1890), pp. 287-304; 22nd Annual Report (1891), pp.

171 C. W. Peckham, “The Growth of Children,” 6th Annual Report of the State
Board of Health of Wisconsin
(1881), pp. 28-73.

182 Franz Boas and Clark Wissler, “Statistics of Growth,” Report of the U. S. Commissioner of Education for 1904 (Washington, 1905), pp. 25-132.

193 W. T. Porter in the Transactions of the Academy of Sciences of St. Louis, “The
Physical Basis of Precocity and Dullness” (1893), pp. 161-181; “The Relation between
the Growth of Children and their Deviation from the Physical Type of Their
Sex and Age” (1893), pp. 263-280; “The Growth of St. Louis Children” (1894), pp.
263-380; also Quarterly Publications of the American Statistical Association, N. S.,
vol. 3 (1893), pp. 577-587; vol. 4 (1894), pp. 28-34.

204 Franz Boas, “The Growth of Toronto Children,” Report of the U.S. Commissioner of Education for 1896-97 (Washington, 1898), pp. 1541-1599.

215 Franz Boas, “The Growth of First-Born Children,” Science, N. S., vol. 1 (1895),
pp. 402-404.

221 Franz Boas, “The Half-Blood Indian, an Anthropometric Study,” Popular Science
, vol. 45 (1894), pp. 761-770 (see pp. 138 et seq. of this volume). Louis
R. Sullivan, “Anthropometry of the Siouan Tribes,” Anthropological Papers of the
American Museum of Natural History
, vol. 23, Part III (1920), p. 199.

232 Roland B. Dixon, Indian Population in the United States and Alaska, 1910
(Washington, 1915),pp. 157-160.

243 “Indian-White Amalgamation,” University of Minnesota Studies in Social Science,
No. 6 (Minneapolis, 1916).

251 See also E. B. Reuter, The Mulatto in the United States (Boston, 1918).

262 “The Physical Growth of Children from Birth to Maturity,” Iowa Child Welfare
Research Station Study
, vol. 1 (1921), No. 1.

271 See footnotes on p. 32.

282 Franz Boas and Helene M. Boas, “The Head Forms of the Italians as Influenced
by Heredity and Environment,” American Anthropologist, N. S., vol. 15 (1913),
pp. 163 et seq.

291 Franz Boas, “Notes on the Anthropology of Sweden,” Amer. Jour. of Physical
, vol. 1 (1918), pp. 415 et seq.

302 Franz Boas, Changes in Bodily Form of Descendants of Immigrants, Columbia
University Press (New York, 1912), pp. 76 et seq.

313 Franz Boas, “Zur Anthropologie der nordamerikanischen Indianer,” Verhandlungen
der Berliner Gesellschaft für Anthropologie, Ethnologie und Urgeschichte
vol. 27 (1895), pp. 404 et seq.

Louis R. Sullivan, “Anthropometry of the Siouan tribes,” Anthropological Papers of the American Museum of Natural History, vol. 23, Part III (1920) pp. 3,
136, 161.