GENETICS: H. J. MULLER 623 
balanced race with another which also contained lethal factors (either 
the same or different in nature and grouping) . In such cases not only 
twin but also multiple hybrids may be produced, that may or may not 
be constant. The results, however, always came out according to the 
prediction based on knowledge of the factorial composition of the flies. 
One such result which was especially noteworthy was a prearranged 
cross in which a dominant character present in one of the parent flies 
was caused to disappear completely, being absent from all the progeny 
of the cross and all subsequent generations. 
(c) A lowered productivity is of course noticeable in balanced races, 
owing to the action of the lethal factors. 
(d) In stocks in which other recessive mutant factors had been in- 
troduced into one or other of the chromosomes containing the lethals, 
these factors, of course, usually failed to manifest themselves, owing 
to the enforced heterozygosis. Occasionally, however, one or more 
crossed over from the lethal factor with which it was bound, and so 
was enabled to become homozygous. As crossing over occurs with pre- 
dictable frequencies, these individuals showing characters abnormal 
to the stock were thrown continually in a definite, very small per cent 
of cases. This caused the stock to appear 'eversporting.' Crossing 
over between the lethal factors themselves also occurred very rarely, 
giving rise to individuals no longer exhibiting the unusual genetic be- 
havior due to balanced lethals. 
6. The striking parallel between the above behavior and that exhib- 
ited in Oenothera make it practically certain that this, too, is a com- 
plicated case of balanced lethal factors, and that some (if not most) of 
the so-called mutations in O. lamarckiana are but the emergence into a 
state of homozygosis, through crossing over, of recessive factors con- 
stantly present in the heterozygous stock. Proof of the spuriousness of 
some of the mutations in Oenothera is, however, not an argument against 
the validity of the modern mutation theory; the fact of real mutation 
has been amply demonstrated in Drosophila as well as elsewhere, and 
it should be emphasized that these mutations can here be distinguished 
with certainty from the superficially similar phenomena that are ob- 
servable in the beaded stock, because the genetic constitution of the 
flies can be analyzed in detail. 
In Oenothera a form of the balanced lethal explanation was suggested 
by de Vries only to account for his double reciprocal crosses, but it is 
evident from the analogy of the beaded case that it probably Hes at the 
root of nearly all the unusual genetic phenomena of the genus. The 
two cases differ in detail, however, in that one or more of the lethals in 
