Epistasis is defined as the influence of the genotype at one locus on
the effect of a mutation at another locus. As such it plays a crucial
role in a variety of evolutionary phenomena such as speciation, popula
tion bottle necks, and the evolution of genetic architecture (i.e., th
e evolution of dominance, canalization, and genetic correlations). In
mathematical population genetics, however, epistasis is often represen
ted as a mere noise term in an additive model of gene effects. In this
paper it is argued that epistasis needs to be scaled in a way that is
more directly related to the mechanisms of evolutionary change. A rev
iew of general measurement theory shows that the scaling of a quantita
tive concept has to reflect the empirical relationships among the obje
cts. To apply these ideas to epistatic mutation effects, it is propose
d to scale A x A epistatic effects as the change in the magnitude of t
he additive effect of a mutation at one locus due to a mutation at a s
econd locus. It is shown that the absolute change in the additive effe
ct at locus A due to a substitution at locus B is always identical to
the absolute change in B due to the substitution at A. The absolute A
x A epistatic effects of A on B and of B on A are identical, even if t
he relative effects can be different. The proposed scaling of A x A ep
istasis leads to particularly simple equations for the decomposition o
f genotypic variance. The Kacser Burns model of metabolic flux is anal
yzed for the presence of epistatic effects on flux. It is shown that t
he non-linearity of the Kacser Burns model is not sufficient to cause
A x A epistasis among the genes coding for enzymes. It is concluded th
at non-linearity of the genotype-phenotype map is not sufficient to ca
use epistasis. Finally, it is shown that there exist correlations amon
g the additive and epistatic effects among pairs of loci, caused by th
e inherent symmetries of Mendelian genetic systems. For instance, it i
s shown that a mutation that has a larger than average additive effect
will tend to decrease the additive effect of a second mutation, i.e.,
it will tend to have a negative (canalizing) interaction with a subse
quent gene substitution. This is confirmed in a preliminary analysis o
f QTL-data for adult body weight in mice.