during the preschool years, which is when psychosexual
differentiation chiefly is accomplished ( Money, Hampson,
& Hampson, 1955, 1957). At this time all development is
susceptible to influence from the social environment--for
example, the acquisition of a native language. That the
differentiation of a gender identity is also powerfully influ-
enced by social experience can be clearly demonstrated in
certain cases of hermaphroditism (see subsequent discus-
sion). As in the case of a native language, however, the
process of psychosexual differentiation is clearly one of
interactionism between brain and social stimulus. The child
with injury to language centers of the dominant hemisphere
may be defective in language acquisition. One cannot be
very specific about the human brain in matters of psycho-
sexual identity, except to note that psychosexual changes
may, in rare instances, be associated with temporal lobe
malfunction or injury; and that these changes sometimes
are reversed as a consequence of successful brain surgery
( Blumer, 1969; Epstein, 1961, 1969). It may also be of
significance that not only psychosexual pathology, but also
electroencephalographic abnormality ( Hambert & Frey,
1964) has an elevated frequency incidence in Klinefelter's
syndrome.

Other syndromes in which the sex chromosomes are im-
plicated are the female triple-X syndrome, Turner's syn-
drome, and the XYY syndrome in males ( Bartalos &
Baramki, 1967; Gardner, 1969; Wilkins, 1965). Triple-X
females are morphologically and psychosexually unremark-
able.

Patients with Turner's syndrome are morphologic females
who are dwarfed, without gonads, and subject to a variety
of other birth defects. The chromosomal error is most
typically an absence of one of the pair of X chromosomes
(45, X). Girls with this condition need hormonal replace-
ment therapy at the age of puberty in order to mature
sexually. Psychosexually they represent virtually the ob-
verse of homosexuality which is conspicuous by its absence
in this syndrome. The girls not only conform to the style of
femininity idealized in our cultural definitions of feminin-
ity, but they also are (long before they know the prognosis
of their condition) maternal in their childhood play and
adult aspirations. This very complete feminine gender iden-
tity and absence of homosexual traits may have its origins
in a total absence of gonadal hormones during fetal devel-
opment, so that there is no malelike hormonal effect on the
sex-regulating centers of the developing brain ( Ehrhardt,
1967; Money & Mittenthal, 1970).

The XYY syndrome (see review by Money, Gaskin, &
Hull, 1970) has so recently been discovered that informa-
tion regarding gender identity in men with this chrom-
osomal aberration (47, XYY) is still tentative. The condi-
tion became noteworthy after it was found to be frequent
among tall and slender men detained in institutions for
delinquents and criminals. Therefore, it has been conjec-
tured that the extra Y chromosome may have some bearing
on poorly regulated or impulsive behavior, sexual behavior
included. The incidence of homosexual experience is high
among institutionalized XYY males and occurs also when
they are not institutionalized.

The examination of syndromes gives one-half of the story
(the half that is traditionally neglected) of the relationship
between genetics, specifically chromosomal genetics, and
homosexuality. The other half of the story comes from
testing a sample group of homosexuals themselves.

Nuclear sex-chromatin surveys of homosexuals and eon-
ists have disclosed no discrepancies between them and con-
trol groups of men with normal masculine gender identity
( Bleuler & Wiedemann, 1956; Pare, 1956; Raboch &
Nedoma, 1958). In chromosome counting, there are no
reports of discrepancies consistently related to homosexual-
ity or to either the transvestite or transsexual form of
eonism ( Pritchard, 1962), though there are known sporadic
combinations of either homosexuality or eonism with the
XXY chromosome complex of Klinefelter's syndrome (as
discussed earlier in this review). Though the total number
of cases studied has been modest, deviations from normal
expectancy have always been in patients whose other clini-
cal signs indicated the probability of a cttigebetuc error in
advance of the actual test. Homosexual men whose physical
examination reveals no bodily abnormality have not been
found to have a Barr body, indicative of an extra X chro-
mosome; nor when the more time-consuming chromosome
count has been performed have chromosomal errors been
directly visualized. Since men with the XYY syndrome
were overlooked until recent years, one must allow the
possibility that a chromosome-counting (karyotyping) sur-
vey of a large sample of homosexuals might disclose some
hitherto unsuspected abnormality in some individuals.
Meantime, on the basis of techniques so far employed,
there is no way of implicating an error of the chromosomes
themselves in the etiology of ordinary homosexuality.
Whether genes rather than entire chromosomes may be
implicated is an altogether different matter. There is no
technique yet available for visualizing, counting, or other-
wise directly implicating certain genes in the etiology of
anything. Such implication must always be by inference, is
only rarely possible, and has not yet been achieved in any
part of behavior genetics, to say nothing of sexual behavior
genetics.

The attempt to implicate hereditary mechanisms in
homosexuality at the genic, if not the chromosomal, level
long antedates the new era of cytogenetics. The older,
statistical methods are those of the sex ratio, ordinal posi-
tion, and twin comparisons.

In sex-ratio studies, the male:female ratio in the sibships
of male homosexuals was compared with the expected ratio
of 106:100 ( Darke, 1948; Jensch, 1941a, 1941b; Kallman,
1952; Lang, 1940; Slater, 1958). Each study turned up a
different ratio, some with and some without statistical
significance, ranging from 106: 100 in Darke's small sample
to 125:100 in Kallman's twin study. The most often
quoted study is that of Lang, based on 1,015 cases. His
ratio was 121:100, which could be subdivided to 128:100
for those cases over the age of 25, and to 113: 100 for the
younger age group. It is possible that the results of all these

-4-