Vol. 6, 1920 
GENETICS: A. WEINSTEIN 
637 
occurring in homologous chromosomes of different species must lie in 
homologous loci. 
A case particularly in point in the present connection is that of the body 
color chlorotic, resembling yellow; its gene is close to the yellow gene in 
the melanogaster X chromosome. Another case is that of magenta in 
virilis. If forked were not known in this species, magenta might have 
been homologized with garnet, which is twelve units to the left of forked 
in melanogaster. While this identification is not entirely impossible, it 
is rendered rather improbable by the difference in distance from forked 
and by the large number of magenta-like mutants in melanogaster. The 
cases of singed and inflated have already been considered. 
Not only may similar effects be caused by non-homologous genes, but 
two originally identical genes may, because of the mutation of one, or the 
mutation of both in different directions, or because of differences in modi- 
fying factors, come to have different somatic effects. Modifying factors 
may also affect the strength of linkage. The differences between cross- 
veinless in melanogaster and virilis, and the differences between the 
lengths of the yellow crossveinless and the crossveinless forked distances 
in the two species, might be thus explained. 
The order of the genes would not be affected by modifying factors 
(though if crossing over were entirely prevented, the order would be im- 
possible to determine, at least in the ordinary way). But Bridges's cases 
of duplication and transposition show that genes may be shifted to 
another region of the chromosome, or even to a different chromosome, 
without being otherwise affected; so that homologous (and indeed ap- 
parently identical) genes may come to have different linkage relations. 
By a combination of mutation and transposition, homologous factors 
might come to differ in both chemical composition and in linkage relations. 
Further complications might be introduced because of duplication and 
of deficiency. In the vermilion duplication stock of melanogaster, each 
X chromosome carries two vermilion genes. Though one vermilion gene 
is recessive to wild-type, two vermilion genes dominate one not- vermilion. 
If a low cross-over factor were introduced, the two vermilion genes 
would remain together and would simulate the behavior of a single 
dominant gene. If vermilion-deficiency were substituted for one ver- 
milion gene, the vermilion vermilion-deficiency combination would simu- 
late a dominant vermilion with a recessive lethal effect. 
A similar reversal of dominance might occur because of the presence 
of several recessive genes in tetraploidy. Such a case could be distin- 
guished by the frequencies of the dift'erent classes, for each factor would 
act as an allelomorph to each of the others. Finally, there might also 
be reversal of dominance because of modifying factors. * 
It must therefore be understood that while the discovery of cross- 
veinless in D. virilis and D. melanogaster increases the probability that 
