Genetic Analysis 285was most explicitly reflected by terms such as “penetrance” and “expressivity”,
the proportion of individuals of a given genotype showing the predicted phenotype
and the degree that it showed, respectively, which were introduced by Vogt [1926].
These were considered to be inherent properties of the genes concerned (or of their
alleles) [Falk, 2000a; Sarkar, 1999]. A major contributor to this genocentricity was
the popularity of the statistical test of the Analysis of Variance that neatly divided
the phenotypic variation into a genotypic and an environmental component (with
a residual “interaction” component) (see Lewontin [1974] for the analysis of the
distortion). As eventually, stated explicitly, the Darwinian notion of natural se-
lection should include the organism shaping its environment as much as that of
the environment shapes the organism [Lewontin, 2000].
Contrary to the early attention to internal and external factors that affect sex-
determination in various species, and despite of intensive studies such as Gold-
schmidt’s of cytoplasmic factors that vary in the “strength” of sex-determination
in the gypsy mothLymantria dispar, Morgan’s students, utilized deviations in
sex determination of chromosomal aberations in Drosophila to establish a genetic-
chromosomally centered theory of sex-determination. According to this the Y-
chromosome did not play a role in sexual differentiation (except of conferring fer-
tility on males) — XXY flies were normal and fertile females, whereas XO flies were
males but sterile. By producing triploid flies with either 3 or 2 X-chromosomes,
Bridges demonstrated that the ratio of X-chromosomes to autosomes (non-sex-
chromosomes) determined the flies’ sex [Bridges, 1925]. For many years this “bal-
ance theory” of sex determination prevailed although exceptions were known, such
as in the bisexual plantMelandrium albawhere the presence of a Y-chromosome
was shown to be needed for male-sex determination [Westergaard, 1958]. It was,
therefore, somewhat of a shock when in the 1950s it turned out that in humans
individuals with XXY chromosomes are phenotypic males (suffering of the Kline-
felter syndrome, that includes sterility) and individuals with XO chromosomes are
females (suffering from Turner’s syndrome).
This notion of an inherent gene action, even when interaction was immanent
may best be illustrated by Muller’s presentation of “position effect”. Whereas
Goldschmidt considered position effect to be evidence for the integrity of the chro-
mosome rather than that of discrete genes [Goldschmidt, 1954], Muller considered
this very property of being able to interact with neighboring genes, as evidence for
the integrity of the discrete genes (see section 5). The advantages of this reduc-
tionist research method were considerable. The most direct attempt to abridge
the experimental-methodological gap between embryology and genetics was that
undertaken by Beadle and Ephrussi [1936], who deliberately applied embryolog-
ical experimental design to genetic problem-posing. Thus they studied the ef-
fects of reciprocal transplantations of imaginal discs (distinct discs of tissue in
the body of the larva, each destined to become a specific adult organ) among
larvae of Drosophila of various eye-color mutants (and wild type) so as to elu-
cidate the developmental pathway of the production of the eye-color pigments,
the ommochromes and the pteridines. In spite of the experimental achievements